The new HPC system, valued at $5.3 million, will support data-driven research in astrophysics, computational biology, health sciences, computational chemistry, materials and manufacturing, and numerous other projects. It will also be used for research that improves the energy efficiency and performance of the HPC systems themselves.
The effort was led by Srinivas Aluru, co-executive director of the Institute for Data Engineering and Science (IDEaS) and professor in the School of Computational Science and Engineering.
“This project is exciting from many perspectives, but especially how it is pushing forward data and high performance computing research infrastructure at Georgia Tech,” said Aluru. “It reflects the teamwork of dozens of faculty, and also supports the work of over 50 research scientists and 200 graduate students.”
Also central to the award are Surya Kalidindi, professor in the George W. Woodruff School of Mechanical Engineering; Charles David Sherrill, professor in the School of Chemistry and Biochemistry; Deirdre Shoemaker, professor in the School of Physics, Rich Vuduc, associate professor in the School of Computational Science and Engineering, and Marilyn Wolf, professor in the School of Electrical and Computer Engineering and the Rhesa "Ray" S. Farmer, Jr. Distinguished Chair in Embedded Computing Systems.
The system is anticipated to begin operations in 2019, and will surpass the current campus capabilities. It will be used for applications that require large memories or local storage, provide modern GPU accelerators, and large storage capacity for data and simulation results.
HPC simulations—one of several uses of the new system—are important for solving large-scale problems in hours or days, rather than months or years. Applications of these include detection of gravitational waves, climate models, performance of materials used in manufacturing or healthcare, and drug discovery.
The new HPC acquisition will coincide with the unveiling of an 80,000 sq.ft. data center in the new Coda building. Coda, the 21-story, 650,000 sq.ft. new addition to Technology Square, lies adjacent to the Georgia Tech campus and major fiber pathways connecting the Southeast.
“We worked to ensure the acquisition is well-timed to be the pivotal supercomputer in the Coda data center,” said Aluru.
“This award is a major boon for interdisciplinary research at Georgia Tech, one that will also be a valuable addition to the HPC-based research community nationally. With Coda opening its doors soon, this supercomputer will become the premier computing resource at Georgia Tech,” said Executive Vice President for Reseach Chaouki Abdallah.
IDEaS and many users of the new equipment will be based in Coda. System management will be handled by the Partnership for an Advanced Computing Environment, or PACE, also residing in Coda.
Research enabled by new system will aid several national initiatives in big data, including strategic computing, materials genome, manufacturing partnerships, NSF-supported observatories such as the LIGO gravitational wave observatory, and the South Pole neutrino observatory known as IceCube.
Researchers from all levels—from early career to undergraduate students—will be the target of training and outreach. Several Georgia Tech researchers and partner institutions will be awarded time on the equipment based on scientific merit and on the national significance of proposed problems.
One-fifth of the system capacity will be dedicated to the research activities of regional partners including minority serving institutions. Other users can participate through XSEDE, a national network of NSF supercomputers that scientists use to interactively share computing resources, data and expertise.
“High performance computing is a priority area for Georgia Tech. Data analysis, simulations, and computational predictive tools are essential elements of modern science, engineering and design. High performance computing is the laboratory of the 21st century,” said Rafael L. Bras, provost and executive vice president for Academic Affairs and K. Harrison Brown Family Chair. “It is extremely satisfying to see a multidisciplinary team successfully work together to make this acquisition a reality. That, after all, is the spirit and culture of Coda.”
]]>
Joshua Chamot
Public Affairs Specialist for Mathematical and Physical Sciences
National Science Foundation
Office of Legislative and Public Affairs
(703) 292-4489
jchamot@nsf.gov
https://twitter.com/NSF_MPS
https://www.facebook.com/US.NSF/
The International Conference for High Performance Computing, Networking, Storage, and Analytics, or “Supercomputing” (SC) for short, was held in Denver from November 12-17 and hosted nearly 14,000 attendees. SC is the premier event for advances in algorithms, applications, architectures and networks, clouds and distributed computing, data analytics and visualization, machine learning and HPC, programming systems, system software, and state of the practice in large-scale data storage, deployment and integration. Each year, SC provides a unique opportunity to meet leaders in the field of high-performance computing, including researchers at universities and government labs, and hardware vendors like Intel, AMD, NVIDIA, and Penguin Computing.
The theme for this year’s event was “I am HPC; Impact and Future Direction,” focusing on the ways in which HPC impacts society as well as the community of researchers in HPC. An interdisciplinary cohort comprised of Georgia Tech researchers from the Partnership for an Advanced Computing Environment, the Center for High-Performance Computing, the School of Computational Science and Engineering, the Center for Research into Novel Computing Hierarchies, the Institute for Data Engineering and Science, and the School of Computer Science at various levels of their career were in attendance to present technical talks, participate in workshops and promote HPC research at Georgia Tech with a booth in the exhibit hall.
Georgia Tech teams were well represented across the research themes, including presentations on; large graph analytics, new methods for high-performance data-intensive workloads, challenges presented by the exascale computing, and Large Language Models (LLMs) applications in codesign. Jeffrey Young and Richard Vuduc were feature speakers in the tutorial “Leveraging SmartNICs for HPC Applications” that offered attendees an in-depth exploration of the state-of-the-art for SmartNICs and the emerging software ecosystems supporting them. Georgia Tech was also represented in the Birds of a Feather Co-Hort Panel Discussions on “Software Testing for Scientific Computing in HPC” and “Scientific Software and the People Who Make It Happen: Building Communities of Practice”. Of special mention is the Special Topic Workshop on “Machine Learning with Graphs in High Performance Computing Environments” organized by Richard Vuduc of Georgia Tech as well as Seung-Hwan Lim, Catherine Schuman, and Jose Moreira.
Georgia Tech researchers had numerous discussions with potential collaborators and new partners for initiatives in high performance computing, including conference attendees from universities, government labs, and industry. The team also had a great opportunity to reconnect with numerous alumni, who stopped by the GT booth to tell us about their careers since graduation. Georgia Tech graduates are doing some amazing things in computing hardware, algorithms, and software, with applications across a wide range of engineering and science problems.
Proposed Workshop: “Foundation of scientific AI (Artificial Intelligence) for Optimization of Complex Systems”
Primary PI: Peng Chen, Assistant Professor, School of Computational Science and Engineering
Proposed Series: “Guest Lecture Seminar Series on Generative Art and Music”
Primary PI: Gil Weinberg, Professor, School of Music
Title: Human-in-the-Loop Musical Audio Source Separation
Topics: Music Informatics, Machine Learning
Primary PI: Alexander Lerch, Associate Professor, School of Music
Co-PIs: Karn Watcharasupat, Music Informatics Group | Yiwei Ding, Music Informatics Group | Pavan Seshadri, Music Informatics Group
Title: Towards A Multi-Species, Multi-Region Foundation Model for Neuroscience
Topics: Data-Centric AI, Neuroscience
Primary PI: Eva Dyer, Assistant Professor, Biomedical Engineering
Title: Multi-point Optimization for Building Sustainable Deep Learning Infrastructure
Topics: Energy Efficient Computing, Deep Learning, AI Systems OPtimization
Primary PI: Divya Mahajan, Assistant Professor, School of Electrical and Computer Engineering, School of Computer Science
Title: Neutrons for Precision Tests of the Standard Model
Topics: Nuclear/Particle Physics, Computational Physics
Primary PI: Aaron Jezghani - OIT-PACE
Title: Continual Pretraining for Egocentric Video
Primary PI: : Zsolt Kira, Assistant Professor, School of Interactive Computing
Co-PI: Shaunak Halbe, Ph.D. Student, Machine Learning
Title: Training More Trustworthy LLMs for Scientific Discovery via Debating and Tool Use
Topics: Trustworthy AI, Large-Language Models, Multi-Agent Systems, AI Optimization
Primary PIs: Chao Zhang, School of Computational Science and Engineering & Bo Dai, College of Computing
Title: Scaling up Foundation AI-based Protein Function Prediction with IDEaS Cyberinfrastructure
Topics: AI, Biology
Primary PI: Yunan Luo, Assistant Professor, School of Computational Science and Engineering
The goals of the funded proposals include identifying prominent emerging research directions on the topic of AI, shaping IDEaS future strategy in the initiative area, building an inclusive and active community of Georgia Tech researchers in the field that potentially include external collaborators, and identifying and preparing groundwork for competing in large-scale grant opportunities in AI and its use in other research fields.
Overview: The goal of this initiative is to bring together expertise in machine learning/AI, high-throughput computing, computational chemistry, and experimental materials synthesis and characterization to accelerate material discovery. Computational chemistry and materials simulations are critical for developing new materials and understanding their behavior and performance, as well as aiding in experimental synthesis and characterization. Machine learning and AI play a pivotal role in accelerating material discovery through data-driven surrogate models, as well as high-throughput and automated synthesis and characterization.
Overview: Zhigang Jiang is currently leading an initiative within IMAT entitled “Quantum responses of topological and magnetic matter” to nurture multi-PI projects. By crosscutting the IMAT initiative with this IDEAS call, we propose to support and feature the applications of AI on predictive and inverse problems in quantum materials. Understanding the limit and capabilities of AI methodologies is a huge barrier of entry for Physics students, because researchers in that field already need heavy training in quantum mechanics, low-temperature physics and chemical synthesis. Our most pressing need is for our AI inclined quantum materials students to find a broader community to engage with and learn. This is the primary problem we aim to solve with this initiative.
Overview: Our objective is to use large language models (LLMs) in conjunction with AI algorithms to identify effective driver proteins, develop screening algorithms that target appropriate binding sites while avoiding deleterious ones, and consider bioavailability and drug resistance factors. LLMs can rapidly analyze vast amounts of information from literature and bioinformatics tools, generating hypotheses and suggesting molecular modifications. By bridging multiple disciplines such as biology, chemistry, and pharmacology, LLMs can provide valuable insights from diverse sources, assisting researchers in making informed decisions. Our aim is to establish a first-in-class, LLM driven research initiative at Georgia Tech that focuses on designing highly effective small molecule libraries to treat challenging diseases. This initiative will go beyond existing AI approaches to molecule generation, which often only consider simple properties like hydrogen bonding or rely on a limited set of proteins to train the LLM and therefore lack generalizability. As a result, this initiative is expected to consistently produce safe and effective disease-specific molecules.
Overview: We are committed to building a team of interdisciplinary & transdisciplinary researchers and practitioners with a shared goal: developing a new framework which model future climatic variations and the interconnected and interdependent energy infrastructure network as complex systems. To achieve this, we will harness the power of cutting-edge climate model outputs, sourced from the Coupled Model Intercomparison Project (CMIP), and integrate approaches from Machine Learning and Deep Learning models. This strategic amalgamation of data and techniques will enable us to gain profound insights into the intricate web of future climate-change-induced extreme weather conditions and their immediate and long-term ramifications on energy infrastructure networks. The seed grant from IDEaS stands as the crucial catalyst for kick-starting this ambitious endeavor. It will empower us to form a collaborative and inclusive community of GT researchers hailing from various domains, including City and Regional Planning, Earth and Atmospheric Science, Computer Science and Electrical Engineering, Civil and Environmental Engineering etc. By drawing upon the wealth of expertise and perspectives from these diverse fields, we aim to foster an environment where innovative ideas and solutions can flourish. In addition to our internal team, we also have plans to collaborate with external partners, including the World Bank, the Stanford Doerr School of Sustainability, and the Berkeley AI Research Initiative, who share our vision of addressing the complex challenges at the intersection of climate and energy infrastructure.
Overview: Our research team envisions a significant trend in the exploration of AI applications for urban flooding hazard forecasting. Georgia Tech possesses a wealth of interdisciplinary expertise, positioning us to make a pioneering contribution to this burgeoning field. We aim to harness the combined strengths of Georgia Tech's experts in civil and environmental engineering, atmospheric and climate science, and data science to chart new territory in this emerging trend. Furthermore, we envision the potential extension of our research efforts towards the development of a real-time hazard forecasting application. This application would incorporate adaptation and mitigation strategies in collaboration with local government agencies, emergency management departments, and researchers in computer engineering and social science studies. Such a holistic approach would address the multifaceted challenges posed by urban flooding. To the best of our knowledge, Georgia Tech currently lacks a dedicated team focused on the fusion of AI and climate/flood research, making this initiative even more pioneering and impactful.
Overview: Most asset management and recycling use technology that has not changed for decades. The use of bar codes and RFID has provided some benefits, such as for retail returns management. Automated sorting of recyclables using magnets, eddy currents, and laser plastics identification has improved municipal recycling. Yet the overall field has been challenged by not-quite-easy-enough identification of products in use or at end of life. AI approaches, including computer vision, data fusion, and machine learning provide the additional capability to make asset management and product recycling easy enough to be nearly autonomous. Georgia Tech is well suited to lead in the development of this application. With its strength in machine learning, robotics, sustainable business, supply chains and logistics, and technology commercialization, Georgia Tech has the multi-disciplinary capability to make this concept a reality, in research and in commercial application.
Overview: Artistic human movement at large, stands at the precipice of a data-driven renaissance. By leveraging novel tools, we can usher in a transparent, data-driven, and accessible training environment. The potential ramifications extend beyond dance. As sports analytics have reshaped our understanding of athletic prowess, a similar approach to dance could redefine our comprehension of human movement, with implications spanning healthcare, construction, rehabilitation, and active aging. Georgia Tech, with its prowess in AI, HCI, and biomechanics is primed to lead this exploration. To actualize this vision, we propose the following research questions with ballet as a prime example of one of the most complex types of artistic movements: 1) What kinds of data - real-time kinematic, kinetic, biomechanical, etc. captured through accessible off-the-shelf technologies, are essential for effective AI assessment in ballet education for young adults?; 2) How can we design and develop an end-to-end ML architecture that assesses artistic and technical performance?; 3) What feedback elements (combination of timing, communication mode, feedback nature, polarity, visualization) are most effective for AI- based dance assessment?; and 4) How does AI-assisted feedback enhance physical wellness, artistic performance, and the learning process in young athletes compared to traditional methods?
Numerous 2022-2023 research findings at Georgia Tech were supported by the NSF-MRI Hive HPC cluster and storage for data-driven discovery, an IDEaS Core Shared Facility. One major project with great local impact was the use of the HIVE to create epidemiological modeling of the COVID -19 outbreak for the State of Georgia. Other finding of note supported by HIVE include advancements in astrophysics related to gravitational waves, intermediate-mass black holes, and the evolution of the universe. Chemistry research results supported include the development simulation created datasets to train Machine Learning models for crystals, hydrogenation catalysis, and negative thermal expansion. The HIVE infrastructure also led to new data-driven analysis & prediction of process-structure-property linkages for rapid new materials discovery.
IDEaS continued to lead in the Data Science world in 2022-2023 as a founding member of the Academic Data Science Alliance (ADSA) and b providing expert voices at U.S. Data Science Leadership Summits. Additionally, IDEaS manages Georgia Tech data science relationships with DOE labs and serves as the lead university in the National Coordination Committee of the Big Data Hubs.
Strategic planning for the 2023-2024 period focused on the rise of AI in government funding decisions and the historical support provided by IDEaS in this area. Specific brainstorming sessions included AI-driven Research in Science and Engineering, Cyberinfrastructure for AI, and Engaging with Industry in the AI Era. Discussions covered ways best pursue future funding and provide support in this area of growing importance for the holistic benefit of Georgia Tech’s different units.
With about 200 affiliate faculty across all six colleges and GTRI, IDEaS is one of the 10 Interdisciplinary Research Institutions under the Georgia Tech EVPR/VPIR. The mission of IDEaS is to support basic and applied research, facilities, training, thought leadership, and external engagement in data science foundations and data driven discovery across disciplines. Learn more at our website.
- Christa M. Ernst
]]>They have developed what is called a ‘cobot,’ which uses table tennis to demonstrate the potential areas a robot can work closely with human partners to complete tasks.
Robot Technology and Ability to Work With Humans
The robot, Barrett WAM arm equipped with a camera and paddle, was trained through a machine learning process called imitation learning. The researchers developed a system to give the robot positive reinforcement for successful volleys, and negative reinforcement for unsuccessful ones.
“We have also trained our robot to be a safe table tennis partner,” said Gombolay. “We leveraged prior work on table tennis and “learning from demonstration techniques” in which a human can teach a robot a skill such as, how to hit a table tennis shot or simply having the human demonstrate the task to the robot."
The project demonstrates the potential for robots to work closely with humans in physical and social capacities, a significant step forward for collaborative robotics, according to Gombolay. The development of intelligent systems that can work collaboratively with humans has numerous applications, from manufacturing, healthcare, to education.
The Work Towards Safe and Trustworthy Human-Robot Interaction
In some cases, researchers found the lack of trust of human participants from explanations given by the robot and were less likely to collaborate with it as a result. A potential reason being a lack of trust that the robot may not have the same goals or motivations as the human partner. Participants in the study were more likely to trust a robot's explanation when they felt that the robot shared their goals and motivations.
“If we can figure out how to safely enable humans and robots to work together in extreme cases, that should give us insights into how to support interaction in a broad variety of settings,” said Gombolay who’s co-study highlights the importance of developing robots that can communicate effectively with humans in a way that builds trust and confidence. This is particularly important in settings where the consequences of a mistake or miscommunication can be severe, such as in healthcare or emergency situations.
“While the choice to work with a robot is ultimately an objective behavior and may vary based upon context or how risky the interaction is, it is ultimately this trust factor that is a key driving force behind your decision-making and behavior,” said Gombolay. “In practice, we often find that people design and deploy impressive robotic solutions, but that robot was not designed to engender the appropriate level of trust from the human end-users.”
Outcome of the Research
The researchers suggest that a possible solution may be to design robots that are more transparent in their decision-making processes. By providing human partners with a clear understanding of how the robot arrived at a particular decision or action, it may be possible to build trust and confidence in the robot's abilities.
“The greater goal of the project is to understand how to design robots for fast-paced, proximate interactions in manufacturing, logistics warehouses, restaurant kitchens, and even in homes. We need to know how to design physically safe systems, of course, but we also need to know what users find intuitive and trustworthy – what makes users feel safe,” said Gombolay. “Otherwise, these robots will never make it out of the lab to coexist with people. I believe our work answers key questions in helping design robots for interaction with people, particularly involving how robots convey their intentions to their human counterparts. But, of course, the research opens even more exciting opportunities than existed before.”
An approach that could eventually be a wave in effective collaboration between humans and robots in a variety of settings.
-Breon Martin
Publication
You can find more information on this project in the recently published paper, “The Effect of Robot Skill Level and Communication in Rapid, Proximate Human-Robot Collaboration,” which include co-authors from the CORE Lab: Kin Man Lee, Arjun Krishna, Zulfiqar Zaidi, Rohan Paleja, Letian Chen, Erin Hedlund-Botti, and Mariah Schrum.
The paper will be featured in the 18th Annual ACM/IEEE International Conference on Human Robot Interaction (HRI), March 13-16, 2023, in Stockholm, SE.
About Matthew Gombolay
Matthew Gombolay is an assistant professor in the School of Interactive Computing at the Georgia Tech. Gombolay is also the director of the Cognitive Optimization and Relational (CORE) Robotics Lab, which seeks to place the power of robots in the hands of everyone by developing new computational methods and human factors insights that enable robots to learn from interaction with diverse, non-expert end-users.
In less than four years, Gombolay’s lab produced more than 50 peer-reviewed publications, including a best paper award at the 2022 ACM/IEEE International Conference on Human-Robot Interaction (HRI), a best paper finalist at the 2020 Conference on Robot Learning (CoRL), and a best student paper finalist at the 2020 American Controls Conference (ACC). Gombolay is a NASA Early Career Fellow and a DARPA Riser and has raised millions in research funding, including support from government agencies (i.e., NSF, NIH, NASA, ONR, and NRL) and industry partners (i.e., Lockheed Martin, Ford, Konica Minolta, and Google) alike. Gombolay has served on Organizing Committees of HRI’21, HRI’22, CoRL’23, and multiple workshops, and he is an Associate Editor of Autonomous Robots and the ACM Transactions on HRI.
1. What is your field of expertise and what questions, or challenges sparked your current research?
My expertise is in quantum chemistry, which is the application of quantum mechanics to problems in chemistry. I had the great fortune to participate in a summer research program in this area as an undergraduate, and I loved its intersection of physics, chemistry, math, and computer simulation. I am especially interested in interactions between molecules, which is relevant for solvation, crystal structures, biomolecular structure, and drug binding. It is an area that is surprisingly hard to model, and requires high-level quantum chemistry techniques.
2. How does the field of Data Science and Engineering intersect with/impact/enhance your research?
I specialize in generating large datasets that can be used to parameterize or test approximate models. The advent of modern machine-learning methods has allowed my group to develop very fast models of intermolecular interactions that are tremendously faster than the quantum chemistry computations that would otherwise be required to achieve a similar accuracy.
3. Why is the field of Data Science and Engineering important to the development of Georgia Tech’s broader research strategy?
Data science, machine learning, and high performance computing have enabled breakthroughs in numerous difficult research areas. They are becoming ubiquitous components of 21st century research.
4. What are the global and social benefits of the research you and your team conduct?
Our work on machine learning models of intermolecular interactions is being used by pharmaceutical companies for improved modeling of drug binding, which will hopefully help speed up the drug discovery process.
5. What are your plans on engaging a wider GT faculty pool with IDEaS research?
Georgia Tech has a strong collaborative spirit, and many faculty in science and engineering whose research would benefit from the latest advances in data science and high performance computing. At the same time, many of those advances are being created by GT researchers in computing. Through IDEaS, I hope to connect more of these researchers and to foster new collaborative efforts.
]]>This week we introduce Xiaoming Huo, A. Russell Chandler III Professor in the H. Milton Stewart School of Industrial and Systems Engineering at Georgia Tech and Associate Director for Research in the Institute for Data Science and Engineering.
Dr. Huo's research interests include statistical theory, statistical computing, and issues related to data analytics. He has made numerous contributions on topics such as sparse representation, wavelets, and statistical problems in detectability.
1. What is your field of expertise and what questions, or challenges sparked your current research?
My research is focused on the foundational principles of data sciences (also known as machine learning, artificial intelligence, etc.). A large spectrum of algorithms has been developed in the data science field, however, there remains a challenge to understand the performance and limitations of these algorithms and then provide helpful, practical guidelines for their usage by researchers. I began my journey to data science as an undergraduate mathematics major. I have been motivated by the emergence of data sciences and the fact that many of these challenging foundational problems are naturally mathematical. What I am doing now seems to be a perfect spot for someone who wants to explore fundamental mathematics while still making an impact in deployable applications.
2. Why is the field of Data Science and Engineering important to the development of Georgia Tech’s broader research strategy?
In the most recent science and engineering research endeavors, no matter the discipline, large amounts of data are generated and collected. Consequently, more and more research finding across colleges and schools at Georgia Tech, and globally, are data-driven. The ability to correlate and make meaningful connections within data will be essential for future researchers. Additionally, the innovative breakthroughs in science and engineering of the future are likely to rely heavily on the utilization of tools from data science and the availability of new data types.
3. What are the global and social benefits of the research you and your team conduct?
Currently, my research is primarily mathematical. Our end goal is to provide data handling principles, guidelines, and best practices that researchers and students can apply.
4. What are your plans on engaging a wider GT faculty pool with IDEaS research?
I am the executive director of the Transdisciplinary Research Institute for Advancing Data Science (TRIAD) (triad.gatech.edu). In 2017, we won an NSF TRIPODS Phase 1 award to establish this Center. TRIAD is housed under the Institute of Data Engineering and Sciences (IDEaS). I am currently serving as the Associate Director for Research (ADR) of IDEaS. As an ADR, I organize campus-wide activities related to data science research at Georgia Tech, aiming to catalyze collaborative activities. My current project is to organize a Georgia Tech workshop on the foundations of data science. We hope to provide a brainstorming event for relevant researchers in early 2023.
Learn More About the Team’s Work Here: https://sites.gatech.edu/xiaoming-huo/
]]>The theme for this year’s event was “HPC accelerates,” focusing on the ways in which discovery in science and engineering is accelerated by high performance computing. An interdisciplinary cohort comprised of Georgia Tech researchers from the Partnership for an Advanced Computing Environment, the Center for High-Performance Computing, the School of Computational Science and Engineering, the Center for Research into Novel Computing Hierarchies, the Institute for Data Engineering and Science, and the School of Computer Science at various levels of their career were in attendance to present technical talks, participate in workshops and promote HPC research at Georgia Tech with a booth in the exhibit hall.
Georgia Tech teams were well represented across the research themes, including presentations on; standardization practices, software engineering, exascale computing, data graphing, and a novel simulator toolkit for co-design. Of special mention is the AMC Gordon Bell Finalist Paper, “Exaflops Biomedical Knowledge Graph Analytics” by Georgia Tech authors Richard Vuduc and Vijay Thakkar.* Georgia Tech faculty were involved in numerous workshops at SC22, including HCP training and education, implementing algorithms on graphics processing units (GPUs), and presenting better tools for developing parallel programs.
Faculty and students from Georgia Tech were integral in the success of Supercomputing 2022, providing planning advice for the organizing committees on Algorithms (Srinivas Aluru), Applications (Umit V. Catalyurek), Architecture and Networks (Tushar Krishna), Data Analytics, Visualization and Storage (Greg Eisenhauer), Machine Learning and HPC (Ramakrishnan Kannan), Post-Moore Computing (Richard Vuduc), Early Career Programs, Student Cluster Competition, and the Student Educational Competitions (Aroua Gharbi).
Georgia Tech researchers had numerous discussions with potential collaborators and new partners for initiatives in high performance computing, including conference attendees from universities, government labs, and industry. The team also had a great opportunity to reconnect with numerous alumni, who stopped by the GT booth to tell us about their careers since graduation. Georgia Tech graduates are doing some amazing things in computing hardware, algorithms, and software, with applications across a wide range of engineering and science problems.
-Christa M. Ernst
*Ramakrishnan Kannan, Piyush Sao, Hao Lu, Jakub Kurzak, Gundolf Schenk, Yongmei Shi, Seung-Hwan Lim, Sharat Israni, Vijay Thakkar, Guojing Cong, Robert Patton, Sergio E. Baranzini, Richard Vuduc, and Thomas Potok. 2022. Exaflops biomedical knowledge graph analytics. In Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis (SC '22). IEEE Press, Article 6, 1–11.
]]>
This week we introduce Regents’ Professor Jeffrey Skolnick, Mary and Maisie Gibson Chair& GRA Eminent Scholar in Computational Systems Biology in the Georgia Tech School of Biological Sciences. Skolnick is also Director of the Center for the Study of Systems Biology and the Thrust Lead for Precision Medicine and Drug Discovery at the Institute for Data Engineering and Science (IDEaS).
1. What is your field of expertise and what questions, or challenges sparked your current research? (Please include at what point in your life you first became interested in this area within your response.)
My field of expertise is computational biology which the goal of understanding the design principles of living systems in general and humans in particular and then to apply this understanding to improve the state of the art of precision medicine and drug discovery. I got into biology by accident. We were doing a series of simulations of polymers and found that the simulations mimicked the properties of proteins. I then realized we could say something original about this important biological system and so I immediately changed fields. Serendipity is very important.
2. How does the field of Data Science and Engineering intersect with/impact/enhance your research?
Learning how cells work relies on big data – what are the states of the literally thousands of different types of molecules (there are 37 trillion cells in a human and each contains millions of molecules). On can then train ML methods to obtain the collective properties of this sea of interacting molecules. Many body interactions are very important determining this behavior- what dictates the cause of a given disease and how can one fix it? Without this field, my research would be impossible.
3. Why is the field of Data Science and Engineering important to the development of Georgia Tech’s broader research strategy?
Data Science and Engineering will ultimately be at the core of all fields of human endeavor. Whether it be weather prediction, health analytics, logistics, or social behavior they do or will rely on the tools of Data Science and Engineering.
4. What are the global and social benefits of the research you and your team conduct?
Our goal is to develop diagnostics and treatments for intractable diseases. These include cancers, e.g. Ovarian cancer which is often diagnosed when it is hard to treat and pain, where we seek to develop non addictive pain relievers. Hopefully our work will help impact these complex problems.
5. What are your plans on engaging a wider GT faculty pool with IDEaS research?
David Sherill and I are planning a workshops to identify those emerging or about to emerge areas where tech can be the leader in the field. We also plan a workshop to identify current opportunities where Tech can play a major role in the advancement of existing fields.
Learn More About the Team’s Work Here: https://sites.gatech.edu/cssb/
]]>Christa M. Ernst - Interdisciplinary Research Communications Program Manager
Topics: Robotics & Data Science
Klaus Advanced Computing Building #1120A
266 Ferst Dr NW, Atlanta | GA 30332
Georgia Institute of Technology | christa.ernst@research.gatech.edu
The four winning teams, led by PIs from across Georgia Tech’s Colleges and Schools, will host their workshops during the Spring 2023 semester at the Georgia Institute of Technology’s Atlanta campus.
Integrative Genomics for Health Equity
This one-day workshop will address the computational and analytical limitations to the use of integrative genomics and multi-omic profiling to understand and promote health equity. As genomic analysis begins to transform healthcare delivery, by promoting personalized assessment of therapeutic intervention, it is becoming increasingly apparent that both social and genetic determinants of health need to be measured. Equitable implementation of genomic medicine must evaluate the influences of ancestry as well as socioeconomic status alongside genetics, with effects mediated in part through gene expression and epigenetics. This workshop will bring together up to 9 speakers who will be asked to present their views on how genomic and non-genomic data can be integrated to guide precision medicine of diverse human populations.
Single Cell Spatial Omics
The field of single cell spatial omics is growing fast, thanks to global consortia such as the Human Cell Atlas (HCA) and the Human Biomolecular Atlas Program (HuBMAP), and also due to reduced next-generation sequencing (NGS) costs. Arguably, the biggest challenge in realizing the full potential of “spatial omics” techniques is the need for analytical tools that maximize our ability to extract testable hypotheses from the rich but noisy data sets. Thus, the AWSOM ’23 workshop will seek to bring together Atlanta-area strengths in computational science and machine learning at the same forum as technology developers and biologists, to strategically determine the thrust areas for future research.
Computational and Mathematical Approaches to Theoretical Neuroscience
Understanding how the human brain works is one of the major challenges of our times. There has been a lot of progress on modeling phenomena at micro scale, such as the model of a neuron, of the chemical channels in a Synapse, learning models for updating weights in neurons etc. Such models have also inspired the models behind modern deep learning architectures. Rapid developments in neuro imaging at both micro and macro levels has enabled us to look at brain phenomena at unprecedented scale. However, an overarching model that explains the macro behavior of the brain is still not found. There have been several exciting steps towards this direction in the last decade from researchers at the intersection of several fields including computational neuroscience, theoretical CS, and probability. The focus of this seminar series is to invite researchers in this space to Georgia Tech, so that students and faculty at GT can pick up and contribute to this young and emerging field.
Sunny Workshop: A Julia Package for The Modeling of Spin Dynamics in Quantum Materials
In recent years, working with scientists at the University of Tennessee and Los Alamos National Laboratory (LANL), we have developed a simulation package called Su(n)ny, that uses Monte- Carlo techniques to calculate the spin dynamics of systems of interests. The Su(n)ny package is written in Julia and currently hosted on Github: https://github.com/SunnySuite/Sunny.jl. Development took place in the last year and a half. Last month, we presented our work for the first time during a workshop at Oak Ridge National Laboratory, as well as tutorials on how to use this package: https://github.com/SunnySuite/SunnyTutorials/tree/main/tutorials, see also the introductory video here: https://mourigal.gatech.edu/public/Sunny-Install-Video-Mourigal.mp4 The package was very well received by our community, and it is now time to accelerate its deployment in realistic community use cases, by coupling it to the modeling of real data, porting it on GPU/Leadership class computers, advertising it more broadly, and including AI/ML methodologies to extract models from data. Learn About the Package Here https://docs.juliahub.com/Sunny/atBCQ/0.3.0/
IDEaS leverages expertise and resources from throughout Georgia Tech's colleges, research labs, and external partners, to define and pursue grand challenges in data science foundations and in data-driven discovery in various fields. For updates on these workshops and other IDEaS events, please visit our website
- Christa M. Ernst
]]>Georgia Tech researchers are presenting their latest work at the annual conference of the North American Chapter of the Association for Computational Linguistics (NAACL 2022), taking place this week, July 10-15. NAACL provides a regional focus for members of the Association for Computational Linguistics (ACL) in North America as well as in Central and South America and promotes cooperation and information exchange among related scientific and professional societies.
“Recent advances in natural language processing – especially around big models – have enabled successful applications,” said Diyi Yang, assistant professor in the School of Interactive Computing and researcher in NLP. “At the same time, we see a growing amount of evidence and concern toward the negative aspects of NLP systems, such as the bias and fragility exhibited by these models, as well as the lack of input from users.”
Yang’s work in computational social science and NLP focuses on how to understand human communication in social context and build socially aware language technologies to support human-to-human and human-computer interaction.
Her SALT Lab has accrued an impressive number of innovations in the field over the past eight months, starting with research presented at last November’s EMNLP conference. SALTers, as they are called, led Georgia Tech to become the top global contributor in computational social science and cultural analytics at that venue. The 60th Meeting of the ACL in Dublin followed in May with multiple SALT studies, including a best paper. Yang’s group has six papers at this week’s NAACL.
“We hope to build NLP systems that are more user centric, more robust, and more aware of human factors,” said Yang. “Our NAACL works are in this direction, covering robustness, toxicity detection, and generalization to new settings.”
Yang’s aspirations for the field are shared by her Tech peers, who have work in the following tracks at NAACL:
Georgia Tech’s research paper acceptances in the main program at NAACL are below. To learn more about NLP and machine learning research at Georgia Tech visit https://ml.gatech.edu.
Ethics, Bias, Fairness
Explaining Toxic Text via Knowledge Enhanced Text Generation
Rohit Sridhar, Diyi Yang
Information Extraction
Self-Training with Differentiable Teacher
Simiao Zuo, Yue Yu, Chen Liang, Haoming Jiang, Siawpeng Er, Chao Zhang, Tuo Zhao, Hongyuan Zha
Information Retrieval
CERES: Pretraining of Graph-Conditioned Transformer for Semi-Structured Session Data
Rui Feng, Chen Luo, Qingyu Yin, Bing Yin, Tuo Zhao, Chao Zhang
Interpretability and Analysis of Models for NLP
Identifying and Mitigating Spurious Correlations for Improving Robustness in NLP Models
Tianlu Wang, Rohit Sridhar, Diyi Yang, Xuezhi Wang
Measure and Improve Robustness in NLP Models: A Survey
Xuezhi Wang, Haohan Wang, Diyi Yang
Reframing Human-AI Collaboration for Generating Free-Text Explanations
Sarah Wiegreffe, Jack Hessel, Swabha Swayamdipta, Mark Riedl, Yejin Choi
Machine Learning
AcTune: Uncertainty-Aware Active Self-Training for Active Fine-Tuning of Pretrained Language Models
Yue Yu, Lingkai Kong, Jieyu Zhang, Rongzhi Zhang, Chao Zhang
MoEBERT: from BERT to Mixture-of-Experts via Importance-Guided Adaptation
Simiao Zuo, Qingru Zhang, Chen Liang, Pengcheng He, Tuo Zhao, Weizhu Chen
Machine Learning for NLP
TreeMix: Compositional Constituency-based Data Augmentation for Natural Language Understanding
Le Zhang, Zichao Yang, Diyi Yang
NLP Applications
Cryptocoin Bubble Detection: A New Dataset, Task & Hyperbolic Models
Ramit Sawhney, Shivam Agarwal, Vivek Mittal, Paolo Rosso, Vikram Nanda, Sudheer Chava
Semantics: Sentence-level Semantics and Textual Inference
SEQZERO: Few-shot Compositional Semantic Parsing with Sequential Prompts and Zero-shot Models
Jingfeng Yang, Haoming Jiang, Qingyu Yin, Danqing Zhang, Bing Yin, Diyi Yang
SUBS: Subtree Substitution for Compositional Semantic Parsing
Jingfeng Yang, Le Zhang, Diyi Yang
Ph.D. student Quintin Kreth, recent Ph.D. graduate Daniel Schiff, and Director of Graduate Research Ethics Jason Borenstein represented the School of Public Policy. They conducted the research in collaboration with Ellen Zegura, professor and Stephen Fleming Chair in Telecommunications in the College of Computing, and Jeonghyun “Jonna” Lee, research scientist in the Center for 21st Century Universities.
In the paper, the researchers note that while scientists and public figures have called for improved ethics and social responsibility education in computer science, there had been no empirical evidence that computer scientists felt less social responsibility than other professionals. The researchers therefore sought to understand how undergraduate computer science students currently view their social responsibilities, especially compared to students in other STEM fields.
After surveying 982 Georgia Tech students from both STEM and non-STEM majors, the research team found that computer science students have “statistically significantly lower social responsibility attitudes than their peers in other science and engineering disciplines.” They argue that these findings suggest a need for additional ethics education in undergraduate computer science programs.
“Computing degree programs have the opportunity to help nurture a mindset within future professionals of sincere interest in protecting the public,” they write. “If not, computing risks diminishing the reputation of the profession even further, increasing regulatory scrutiny, and exposing the public to greater harms.”
In giving the Georgia Tech team its awards, the reviewers praised the researchers for their clear writing and analysis, as well as the timeliness of their topic.
“It was a great honor and surprise to receive this award,” said Kreth. “I’m told student-led papers rarely win it.”
Because they were a finalist for best overall paper, the authors have been invited to present their work again next year, at the 2023 ASEE conference, as part of a showcase event.
The paper is forthcoming and will be available online in the coming weeks at https://peer.asee.org/.
]]>Abernethy, based in the School of Computer Science, and his Emory partner, anthropology assistant professor Marcela Benítez, will use the seed grant to fund their proposal AI Forest: Cognition in the Wild. The pair plans to create and deploy “smart” testing stations for long-term cognitive assessment and monitoring of wild capuchin monkeys at the Taboga Forest Reserve in Costa Rica.
The testing stations will leverage AI and deep learning to recognize and track monkeys in real-time. This means individuals can be targeted for behavioral assessment and cognitive testing. The stations will also provide a new approach to long-term monitoring of cognitive abilities in wild animals, which will provide the team with an unprecedented level of control in a wild environment. This in turn will provide new opportunities for several studies linking cognitive performance to natural behaviors and ultimately overall fitness.
Arriaga is based in the School of Interactive Computing. She and her partner, Dr. Marcos Schechter from the Emory School of Medicine, will use the seed funding to develop a diabetic ulcer computational sensing system. The goal is to build computational models that can detect changes in diabetic foot ulcers and provide AI-driven interfaces that connect patients and clinicians.
To address the current clinical challenges of automating wound screening and monitoring, the human-centered computational sensing system proposed by the team will characterize ulcer severity and wound progression, and predict wound healing and recurrence.
Additionally, the team will focus on underserved and minority communities to promote technologies to reduce disparities. This pilot proposal will enroll people from underserved communities at Grady Memorial Hospital, a public hospital where over 250 people are hospitalized with diabetic foot ulcers annually.
“These teams are partnering to revolutionize AI and promote equity and improvement of the overall quality of human life,” said Deborah Bruner, senior vice president for Research at Emory University. “This is an exciting time for research departments at Emory and Georgia Tech.”
"This opportunity leverages our universities' combined strengths in AI, health, and cognition research,” said Rob Butera, Georgia Tech vice president for Research Development and Operations. “Emory and Georgia Tech have a long history in collaborating and AI.Humanity pushes these collaborations into new domains."
A third team research pair also received AI.Humanity funding for their proposal.
John Taylor, School of Civil and Environmental Engineering Frederick Law Olmsted Professor, and Emory public health professor Lance Waller will use the seed grant for their project Applying Machine Learning Techniques to Improve Epidemiological Models Accounting for Urban Infrastructure Networks, Human Behavioral Change, and Policy Interventions.
]]>Presented at the 14th Annual ACM Web Science Conference, being held this week in Barcelona, Spain, the award recognizes the lasting impact of the team’s 2013 paper Social Media as a Measurement Tool of Depression in Populations.
“When we wrote this paper, little did we know what it meant for times to come. We are energized more than ever before to see how this work can push the boundaries of population mental health,” De Choudhury, an associate professor in the School of Interactive Computing, said in a tweet on June 26.
The award was announced during the conference’s opening session. Dame Wendy Hall, Web Science Trust executive director, made the announcement saying that it was difficult at first to choose from among all the groundbreaking papers presented over the last decade at the Web Science Conference.
“But the trustees felt that this paper captured an important analysis of nearly 70,000 user posts to develop a novel social media depression index—helping us to understand more about the society that we shape (and that shapes us) through the Web.
“The paper has been well-cited, and the issue of mental health certainly retains much relevance today,” said Hall.
The paper, written by De Choudhury, Scott Counts, and Eric Horvitz, was first presented at the 2013 ACM Web Science Conference in Paris, France. As part of the recognition of the work’s impact, De Choudhury is re-presenting the award-winning paper, and discussing its impact, during an online keynote presentation, June 29, at 9:45 a.m. EDT (3:45 p.m. CEST).
The Web Science Trust is a UK charity promoting the understanding of the web, through education and research in the discipline of web science.
]]>The award recognizes a senior member of the Institute of Industrial and Systems Engineers (IISE) community with impactful contributions in innovation and service to industrial engineering. No more than 20 fellows can be named each year. IISE is the world’s largest professional society dedicated to the support of industrial and systems engineers.
Gabraeel’s research interests lie at the intersection of predictive analytics and machine learning, repair and operations, and service logistics. His key focus is developing fundamental statistical learning algorithms specifically tailored for real-time equipment diagnostics and prognostics, and optimization models for subsequent operational and logistical decision-making in Internet of Things ecosystems. From the standpoint of application domains, his general interests lie in manufacturing, power generation, and service-type industries.
Gabraeel leads the Predictive Analytics and Intelligent Systems research group in Georgia Tech’s Supply Chain and Logistics Institute.
]]>Sasha Boldyreva’s paper that her former Ph.D. student, Adam O’Neill, presented at the Crypto 2007 conference has been cited over 1,000 times according to Google Scholar and is one of the first papers to formally study the security of efficiently searchable encryptions for cloud storage.
When Boldyreva first came to Georgia Tech she talked to database researchers and learned about the problems they had securing remote servers. Specifically, faculty were not able to search and retrieve encrypted data while also being able to maintain security.
“You want to keep your data secure, but also need to be able to retrieve it,” Boldyreva said. “I was constantly getting asked if there was a way to make searchable encryption more efficient.”
The standard method of encryption randomizes data to maintain security. However, database researchers found when they retrieved information from a remote server it was often scrambled in a way that couldn’t be deciphered.
To work around this, database faculty used deterministic encryption as a limited solution to search and retrieve parts of remote data without losing any information in the process. This made the process easier, but much less secure. In fact, because the coded text is always the same for a given plain text message, studying the application of deterministic encryption was ignored by cryptographers at the time due to its weak security.
“What is a good deterministic scheme? No one really knew,” Boldyreva said. “My Ph.D. student, my former Ph.D. advisor, and I decided to study the bigger problem of efficient searchable encryption.”
The team began by formally defined the term “efficiently searchable encryption” and proposing specific cryptographic schemes to make deterministic encryption more secure. While the team’s results showed deterministic encryption still posed security risks, Boldyreva said the goal was to define a starting point and allow academics to build from there.
“This has become a very hot topic in cryptography,” she said. “Researchers today are targeting searchable encryption with much stronger security guarantees, but they still view our paper as a starting point.”
Since the paper published in 2007, Boldyreva has continued working on searchable encryption. She recently built one of the most secure searchable encryption schemes to date with the help of her current Ph.D. students.
“This has remained my favorite area of cryptography,” she said. “It is nice to receive recognition for the challenging work we put into our research.”
The award will be presented at the Crypto 2022 conference in August. This is Boldyreva’s second test of time award. She was also recognized in 2020 by the International Conference on Practice and Theory in Public Key Cryptography (PKC) for her work on new multi-user digital signatures in 2003.
Deterministic and Efficiently Searchable Encryption was written by Boldyreva, who currently serves as the associate chair of graduate studies at the School of Cybersecurity and Privacy, O'Neill, and Mihir Bellare, her former faculty advisor.
]]>To help solve this problem, Yuxi Wu, a Ph.D. student at the Georgia Tech School of Interactive Computing (School of IC) explored designing a system that guides people affected by institutional privacy violations toward making unified demands for redress. Wu will present the research at the ACM CHI Conference on Human Factors in Computing Systems (CHI 2022) on May 2.
The paper investigates issues of representation and stewardship that appear when a collective wants change. Wu and her co-authors note that in the cases reviewed by their team, advocacy efforts failed due to a lack of debate and agreed solutions in internet petitions. Unclear goals and a lack of technical knowledge creates a rift between those affected by data breaches and the experts who can create solutions.
“Tons of people go on the internet to complain about these breaches and maybe a few petitions go truly viral,” Wu said. “But the original authors weren’t thinking of thousands of people when they wrote them. There is no good way to have a true discussion and present a united front as a collective.”
For example, in the wake of the Equifax breach in 2017, several petitions were started online. One of the biggest has close to 250,000 signatures with more being added, but it hasn’t been resolved.
The petition in this case states once it has reached 300,000 signatures it will be sent to several federal organizations including a federal judge who ruled on the case in the summer of June 2021. It is also worth noting the creator of the Change.org petition has not made an update in three years.
To explore how to improve on the dead ends of petitions, Wu and her co-authors developed their own version of a find-fix-verify process to corral people toward consensus. They recruited 400 online participants to identify concerns in response to an institutional privacy violation, come up with demands from these concerns, and vote on the most pressing demands. The participants eventually voted for 12 demands.
“People are really angry with big institutions, and they want to talk about them,” said Wu. “It is really difficult to translate these broad concerns and demands into specific actionable priorities.”
To help with mapping these demands onto the real world, the researchers then consulted a panel of eight security and privacy experts.
“The experts did share some agreements with the crowd: they recognized the harms that users faced was painful and hard to deal with,” said Wu. “However, they disagreed with the actions being demanded of institutions.”
A major disconnect between the two groups came from experts deeming the demands as unrealistic. However, Wu points out that experts’ deep knowledge of and entrenchment in the very institutions that collectives are fighting against might also make them dismissive of collectives. As a result, she hopes that future design work can examine how to address this tension, and how experts might in turn be assessed by collectives.
“A Reasonable Thing to Ask For”: Towards a Unified Voice in Privacy Collective Action was written by Wu along with her academic advisors Keith Edwards and Sauvik Das.
]]>Once matched, the mentors provide feedback on their applicant’s CV and statement of purpose. Mentors also share their experiences to provide better insight into the application process for prospective students. The feedback is provided in writing and may include virtual one-on-one meetings as well.
According to program founder Shirley Anugrah Hayati, a first-year HCC Ph.D. student, several applicant/mentor pairs were matched before the Nov. 15 GT-GAS deadline. The applicants will receive feedback in time for them to meet the upcoming Ph.D. application deadline on Dec. 15.
Hayati says that similar programs exist at other institutions, so she decided to develop this service for Georgia Tech. To help organize an assistance program in the College of Computing, she recruited machine learning Ph.D. student Will Held.
“As a minority in computer science, I can imagine that if I didn’t have mentors or anyone to help with my application, how hard that would have been, so I’m inspired to help underrepresented students in applying for computer science-related Ph.D. programs at Georgia Tech,” said Hayati.
After consulting their faculty advisor, Diyi Yang, assistant professor in the School of Interactive Computing, the pair moved forward with the project by developing survey forms for potential applicants and mentors and establishing an implementation timeline for the program.
“GT-GAS is a wonderful initiative by our students that should increase the diversity of graduate education at Georgia Tech. More than 30 Ph.D.’s from the College have joined the program. I am sure there will be more joining next year, and I believe GT-GAS is well on the way to helping more who are in need,” said Yang.
With a successful launch this semester, Hayati hopes to be able to reach more underrepresented applicants in the future and possibly expand the program across campus. As a first step though, she’d like to see the program added to the College website.
“I think bringing this to the College’s official website will help people to find it even earlier so that the applicants have even more time to prepare their applications,” said Hayati.
According to Ann Claycombe, director of communications for the College of Computing, a page with links and information about the GT-GAS program will be added to the College’s newly revamped website next semester.
]]>“Microsoft has been a consistent and generous supporter of IDEaS since its inception, contributing funding for research and events, and providing cloud resources for research and education at Georgia Tech as well as for supporting programs under the South Big Data Hub,” said Srinivas Aluru, executive director of IDEaS. “The launch of the Cloud Hub takes this relationship to the next level and reflects our ongoing commitment to partnership.”
Umakishore Ramachandran, IDEaS thrust lead for cloud computing, will manage the Cloud Hub for IDEaS. Under Ramachandran’s guidance, faculty and researchers can request resources and further Cloud Hub-based projects.
“Cloud computing is a disruptive technology that is powering many aspects of our everyday lives,” said Ramachandran. “Research at Georgia Tech already touches various aspects of Cloud computing spanning systems, machine learning, and data analytics. Cloud Hub launched with Microsoft partnership, facilitates access to Cloud resources for education and research, ensuring that our students will graduate with immersion into this important technology.”
The Hub will utilize Microsoft Azure, a growing collection of integrated cloud services that researchers and educators use to build, deploy, and manage applications through the global network of data centers. Azure provides the freedom to build and deploy wherever the user wants, using the tools, applications, and frameworks they choose.
"The Cloud Hub represents an exciting new direction in how we use computing infrastructure to advance research and education at Georgia Tech,” said Vivek Sarkar, chair, School of Computer Science, professor and Stephen Fleming Chair for Telecommunications. “Industry partnership is critical to growing this direction on campus, and we are grateful to Microsoft for their partnership to get us started.”
Not only will Microsoft help move the Hub in the right direction, their continuous innovation will also ensure Tech faculty and researchers are future-ready.
“We are thrilled to partner with Georgia Tech at this capacity,” said Yousef Khalidi, corporate vice president, Azure for Operators, Microsoft. “The Cloud Hub will provide Georgia Tech faculty and researchers cloud and edge computing technology and solutions, allowing them to realize new opportunities.”
]]>After a campuswide search, Xiaoming Huo is joining IDEaS as Associate Director for Research. New Thrust Lead positions were created positions to focus on and opportunistically expand capabilities in important areas. Joining IDEaS in this capacity are Jeffrey Skolnick as Thrust Lead for Precision Medicine and Drug Discovery and Umakishore Ramachandran as Thrust Lead for Cloud Computing.
In addition, IDEaS is increasingly being called upon to support data and cyber infrastructure technical design and management needs of large, center-scale projects. To adress these needs, senior research scientist Tony Pan will be assuming the role of Assistant Director for Data Infrastructure. David Sherrill, who is already serving as Associate Director for Research and Education, is taking on additional responsibility as Director of the Center for High Performance Computing (CHiPC).
Xiaoming Huo, Associate Director for Research
Xiaoming Huo is an A. Russell Chandler III Professor in the H. Milton Stewart School of Industrial and Systems Engineering at Georgia Tech. Dr. Huo's research interests include statistical theory, statistical computing, and issues related to data analytics. He has made numerous contributions on topics such as sparse representation, wavelets, and statistical problems in detectability. His papers appeared in top journals, and some of them are highly cited. He is a senior member of IEEE since May 2004.
In this new role, Huo brings experience in creating teams to tackle various challenges in science and society. He believes that nurturing and fusing teams within IDEaS will results in more funding opportunities, experience sharing, shaping future programs at the national level, and enhancing the visibility of IDEaS and Georgia Tech.
Jeffrey Skolnick, Thrust Lead for Precision Medicine and Drug Discovery
Jeffrey Skolnick, Regent’s Professor of Biological Sciences, research has focused
on health and life sciences, having developed novel AI approaches for precision medicine, disease mode of action prediction, and drug efficacy and side effect prediction that is at the state of-the-art. He has developed and applied algorithms to proteomes for the prediction of protein structure and function, the prediction of small molecule ligand-protein interactions with applications to drug discovery and the prediction of off-target uses of existing drugs with applications to aging, cancer and chronic fatigue syndrome, cancer metabolomics, precision medicine, fundamental studies on the nature and completeness of protein structure space and the exploration of the interplay between protein physics and evolution in determining protein structure and function, prediction of protein-protein and protein-DNA interactions, and molecular simulations of subcellular processes.
Skolnick hopes to catalyze the development of novel big data-based approaches to Precision Medicine and Drug discovery. In particular, he hopes to identify and catalyze teams that would to transformation research in these areas. Some representative projects include Cancer Multi-omics where the goal would be to stratify patients to predict which patients are likely to respond to specific drugs. Ideally, this would push the development of cancer therapeutics to treat currently intractable cancers such as pancreatic and triple negative breast cancer. Another area is neuroscience, with particular emphasis on Alzheimer’s Disease and Parkinson’s Disease, which currently lack effective, long term treatments. Here, the goal is to identify patient specific disease drivers and key mode of action proteins and non-coding regions responsible for disease onset and progression and then identify, and in collaboration with Emory, test predicted novel repurposed drugs in patients. The goal would also be to create a knowledgebase and website which would make the resulting tools widely available.
Umakishore Ramachandran, Thrust Lead for Cloud Computing
Kishore Ramachandran received his Ph.D. in Computer Science from the University of Wisconsin, Madison in 1986, and has been on the faculty of Georgia Tech since then. He led the definition of the curriculum and the implementation for an online MS program in Computer Science (OMSCS) using MOOC technology for the College of Computing, which is currently providing an opportunity for students world-wide (with an enrollment of over 10,000) to pursue a low-cost graduate education in computer science. He has served as the Director of STAR Center from 2007 to 2014, and as the Director of Korean Programs for the College of Computing from 2007 to 2011. Ramachandran has also served as the Chair of the Core Computing Division within the College of Computing. His research interests are in architectural design, programming, and analysis of parallel and distributed systems. Currently, he is leading a project that deals with large-scale situation awareness using distributed camera networks and multi-modal sensing with applications to surveillance, connected vehicles, and transportation. He is the recipient of an NSF PYI Award in 1990, the Georgia Tech doctoral thesis advisor award in 1993, the College of Computing Outstanding Senior Research Faculty award in 1996, the College of Computing Dean's Award in 2003 and 2014, the College of Computing William "Gus'' Baird Teaching Award in 2004, the "Peter A. Freeman Faculty Award" from the College of Computing in 2009 and in 2013, the Outstanding Faculty Mentor Award from the College of Computing in 2014, and became an IEEE Fellow in 2014.
Ramachandran will help establish a “Cloud Hub” at Georgia Tech partnering with Microsoft. The Cloud hub will provide Microsoft Azure resources for both education and research for faculty and students wishing to use the Cloud for their computational needs. While the Cloud hub will initially be started with support from Microsoft, the aspirational goals for the hub include expansion to include other Cloud providers who may want to partner with Georgia Tech. Further, Ramachandran hopes to use the experience with the Cloud hub to help evolve a Cloud strategy for Georgia Tech as a whole.
Tony Pan, Assistant Director for Data Infrastructure
Tony Pan joined the Institute for Data Engineering and Science in 2018 after graduating from Georgia Institute of Technology with a Ph.D. in Computational Science and Engineering. For over two decades, Pan has focused his research efforts on developing data science methods to enable large scale biomedical and bioinformatic studies, specifically through flexible and extensible data management, high performance computing (HPC) approaches, and efficient parallel algorithms. He is leading the data management infrastructure definition and implementation to support the NSF Engineering Research Center for Cell Manufacturing Technologies (CMaT) at Georgia Tech, and developing HPC algorithms for high-throughput sequencing data, gene networks, and single cell sequence analysis. Pan is also leading the development of data management infrastructure and gene association studies for the Arthrogryposis Registry in collaboration with Shriner’s Hospitals for Children.
In this role, Pan hopes to engage and support IDEaS members, partners, and collaborators in large scale, systematic data management and analysis efforts. Towards this end, he will focus on defining and implementing an IDEaS core data management strategy and corresponding infrastructure, as well as developing common optimization approaches and algorithms for large scale data analytics. Pan hope to foster an environment for data, knowledge, and best practice sharing between researchers, collaborators, and institutional support as part of IDEaS' data infrastructure efforts.
David Sherrill, Director, Center for High Performance Computing
Professor David Sherrill is the new Director for the Center for High Performance Computing (CHiPC). Sherrill is currently the assistant director for research and education in IDEaS. Sherrill is one of the co-PI's on the NSF MRI grant that funded the Georgia Tech Hive computer, and recently organized the Hive Supercomputer Symposium hosted by IDEaS. His background is in theoretical chemistry, and he develop new models in quantum chemistry, with a particular focus on biophysics, drug docking, and molecular crystals. Sherrill's group makes heavy use of high performance computing (HPC) resources, especially now, pioneering how to apply machine learning (ML) to intermolecular interactions. They are creating quantum chemistry datasets of unprecedented size to train these ML models. The existing quantum chemistry software is too slow for this, so they have developed their own very popular open-source program, Psi4.
In this new role, Sherrill hopes to strengthen connections between HPC users in science and engineering with HPC researchers in computing. One goal is to ensure Georgia Tech is prepared to respond to new national initiatives in which HPC may be a key component. Maintaining significant local HPC resources is a key part of that, and CHiPC should facilitate the organization of teams to compete for equipment grants. Sherrill plans to set up education and outreach workshops, and continue the excellent CHiPC efforts related to the student cluster competition and the Georgia Tech booth at Supercomputing.
]]>
ICML is the leading international academic conference in machine learning. Along with NeurIPS and ICLR, it is one of the three primary conferences of high impact in machine learning and artificial intelligence research. It is supported by the International Machine Learning Society (IMLS).
Explore Georgia Tech people, research abstracts, and when authors will present (Tues-Thurs) in an interactive data graphic of Georgia Tech at IMCL 2021. Also explore the whole program in a second data graphic: Who’s Who at ICML 2021.
Georgia Tech’s work is represented in 2% of the program with 22 papers in a range of topics including (asterisk denotes a single paper):
The study was conducted collaboratively by researchers at the Georgia Institute of Technology, the University of North Carolina (UNC), and North Carolina State University (NCSU) and the findings published in the research journal JAMA Network Open. The study methods were based on a mathematical simulation originally developed at Georgia Tech.
The study evaluated how many Covid-19 cases could be avoided in the Tar Heel State if more people get vaccinated and continue to follow mask and physical distancing guidelines. As of June 3, North Carolina has had 1 million reported cases of Covid-19 and more than 13,000 recorded deaths.
“The main takeaway from the paper is that while the increasing vaccine coverage in the U.S. has a positive impact, we are not really there yet. We still need to follow preventive measures such as mask wearing,” said contributing author Pinar Keskinocak, the William W. George Chair and Professor in the H. Milton Stewart School of Industrial and Systems Engineering at Georgia Tech and co-founder and director of the Center for Health and Humanitarian Systems.
The caution is well founded when the researchers account for viral mutations, including the variant currently dominant in the United States that was initially identified in the UK and was associated with the surge in Michigan. There, as recently as May 2, the state averaged nearly 3,500 cases a day, according to a June 2 story in Bridge Michigan.
According to one scenario from the simulation, which was populated with data from the state of North Carolina, if 75% of the population gets fully vaccinated but continues to wear masks and socially distance, there is a sustained decline down to very few new Covid-19 cases over a six-month period. But, if only 25% of the population gets fully vaccinated and does not adhere to these non-pharmaceutical interventions (NPIs), there could be a steady increase in daily Covid-19 cases, peaking around 8,000 before there is another decline.
Keskinocak points to the Georgia Covid-19 Vaccination Dashboard that tracks county-level differences in vaccination rates based on race as further evidence of the need for caution.
“Our dashboard shows that there has been high variability in the level of vaccination in different geographic regions and communities,” she said. “So even if we say over half of U.S. adults are vaccinated, it’s not uniform across the entire country. This further raises concerns about quickly lifting the NPIs.”
Julie Swann, department head, North Carolina State University’s Fitts Department of Industrial and Systems Engineering, and long-time research collaborator with Keskinocak, concurs. “Current variants are more infectious, and there are still locations with less than 30% of the population vaccinated.”
Swann adds that ongoing spread “endangers people now. It also increases the chance of a future mutation that could increase the risk even to people who are vaccinated.”
Swann and Keskinocak are two of three researchers who co-founded Georgia Tech’s Center for Health and Humanitarian Systems in 2007, the year they developed a comprehensive, agent-based simulation model for pandemic flu, which has since been at the core of their modeling efforts for Covid-19. Over the years the two industrial engineers have collaborated closely to advance the model and make results available to decision makers as new pandemics emerged.
“Julie and I have been working on infectious disease modeling for over a decade now. We had developed this agent-based model for pandemic flu and then when Covid-19 hit, Georgia Tech adapted that model to Covid-19 and shared it with Julie’s team at NC State who then modified the model to test additional scenarios and calibrated it with North Carolina data,” Keskinocak said.
“I am grateful that Pinar and I had spent such a long time trying to understand pandemics and improving the science behind them,” Swann added. “If we had not invested that time, we would not be able to have such a fast turnaround and the high participation level that we have this year.”
According to Keskinocak, the simulation model is extremely detailed – in essence, it recreated demographics of the population down to the state’s household size, and even individuals’ workflow, to give a clear picture of how people move from one geographic area to another.
“The model is flexible and can change based on the research question being asked, including the current research question, ‘What is the impact of lifting NPIs in increasing vaccine coverage?’”
Swann credited the close partnership with public health partners in Georgia and North Carolina, as well as their work under a grant funded by the Centers for Disease Control and Prevention and the Council for State and Territorial Epidemiologists, with the speed of developing models to test interventions during the pandemic.
“It’s really key because sometimes we get additional ideas or research questions from the stakeholders with whom we interact,” Swann said.
Both researchers emphasized the critical role students played in advancing these models under tight deadlines while juggling coursework. Between them they estimate more than two dozen graduate students across the partnering institutions have been engaged in the Covid-19 modeling work since the pandemic began. The two also were integral in establishing a professional education HHSCM certificate program.
# # #
The Georgia Institute of Technology, or Georgia Tech, is a top 10 public research university developing leaders who advance technology and improve the human condition.
The Institute offers business, computing, design, engineering, liberal arts, and sciences degrees. Its nearly 40,000 students, representing 50 states and 149 countries, study at the main campus in Atlanta, at campuses in France and China, and through distance and online learning.
As a leading technological university, Georgia Tech is an engine of economic development for Georgia, the Southeast, and the nation, conducting more than $1 billion in research annually for government, industry, and society.
Writer: Anne Wainscott-Sargent
]]>Developed by School of Computational Science and Engineering (CSE) Associate Professor Polo Chau and Prairie View A&M University Assistant Professor Xishuang Dong, a teaching kit focused on data science education as part of the NVIDIA Deep Learning Institute (DLI) is now available for free to qualified educators.
Released as a multipart series, four kits are offered through NVIDIA’s DLI program in collaboration with leading researchers and professors in four research areas:
Specifically, each of these four DLI Teaching Kits lowers the barrier to entry for educators seeking to incorporate artificial intelligence (AI) and graphic processing unit (GPU) computing in coursework by providing downloadable teaching materials and online courses.
The kit teaches students fundamental and advanced topics on accelerated data science with the NVIDIA RAPIDS framework, GPU-accelerated machine learning, data visualization, graph analytics, and more.
“Traditional data science software libraries are mainly written for CPUs and don’t take advantage of GPUs. The RAPIDS library is NVIDIA’s effort to simplify and more easily use their GPU Python focused library,” said Chau.
The Data Science Teaching Kit contains tens of modules and labs, with content adapted from the popular course, CSE 6242: Data and Visual Analytics, taught by Chau. Georgia Tech contributors on the project include Polo Club for Data Science researchers Scott Freitas, Haekyu Park, Jay Want, Jon Saad-Falcon, Kevin Li, Aiswarya Bhagavatula, and Frank Zhou.
“The new development on our side is creating the modules and figuring out how to provide interactive labs for the students to work on and new coding questions,” said Freitas, a lead researcher on the project. “We also released three papers and each one of those papers will inform a lab in the teaching kit.”
Part of these papers include two new large-scale datasets for cybersecurity which are incorporated into the toolkit to teach participants how to detect malware using new graph techniques.
According to Freitas, the two datasets being integrated are also two of the largest cybersecurity and graph datasets ever released in the world.
“The end goal is to help people learn how to use new state-of-the-art GPU accelerated techniques. NVIDIA has many advanced technologies that they are developing but it may not necessarily be accessible to people just getting into the field. So, this teaching kit aims to take all of these components and simplify them in a way that is successful and easy to use for educators,” he said.
]]>
Communications Officer
]]>The data science community and members of the South Big Data Hub should mark their calendars for the very first Southern Data Science Conference, to be held on April 7 at the Hyatt Regency Atlanta Perimeter at Villa Christina. The conference is expected to attract data science thought leaders from around the southeast and the nation and will feature speakers from innovative companies and research laboratories, such as Google, Microsoft, AT&T, NASA, Glassdoor and Groupon.
Those who attend the conference will learn about data science best practices and cutting-edge technologies that are transforming data science and analytics into a game-changing domain that spans business, academia and government. Conference speakers will cover topics ranging from data science fundamentals to building hands-on data science solutions to solve real-world problems in areas such as fraud detection, cybersecurity, relevant search and recommendations, and many others.
The conference also will be the first to host three authors of respected books on data science: Joel Grus author of Data Science from Scratch; Michael Brzustowicz, author of Data Science with Java; and Trey Grainger author of Solr in Action.
Registering early will translate into money saved. The conference offers a $75 early bird discount to those who register by Feb. 15 and will provide an additional 15 percent discount to South Big Data Hub members when they use the discount code SBDHUB2017. Organizations that register a group of five or more attendees will be eligible for a company pass that provides a 20 percent discount.
Conference sponsors include Career Builder, LexisNexis, Datacamp.com, the Becoming a Data Scientist podcast and blog (podcast sponsor), PyData Atlanta, Applied Data Science Atlanta, and the Institute for Data Engineering and Science at Georgia Tech. Sponsorship opportunities are still available. Interested sponsors should contact Khalifeh AlJadda at khalifefh@southerndatascience.com.
To learn more about the conference or to register, please visit the conference website.
]]>The awards are expected to average about $5,000. We remain flexible regarding the proposed use of funds and the total budget amount. The call is open to all Institute faculty regardless of their affiliation with IDEaS. The overarching goal of this effort is to permit sensible planning, provide funding quickly to facilitate early group formation, and increase chances of success for Georgia Tech in this important national competition. IDEaS will also be ready to provide organizational and logistics support for brainstorming retreats and/or proposal writing workshops, as needed.
Interested PIs are requested to submit a one-page proposal detailing:
Proposals should be submitted in a PDF format to Shkina Halbert (shalbert@gatech.edu) by October 31, 2019. Funding decisions will be communicated shortly thereafter. We will continue to entertain proposals as they come in after the October 31 deadline.
]]>IDEaS is one of Georgia Tech's 11 interdisciplinary research institutes within the Georgia Tech Research enterprise.
What is your field of expertise and why did you choose it?
As a graduate student in the early 90s, I chose high performance computing for the same thrill people experience when driving sports cars. There are these wonderful, expensive, and high-performance supercomputers readily available, and writing superfast code and solving the largest-scale problems is such a rush!
My core areas of expertise are high performance computing and data science. Using these as a launching pad, I work in large-scale applications in science and engineering. For the past 25 years, I have done so much work in bioinformatics and computational biology that it has become another core area.
What makes the way in which your IRI enables campus research unique?
Data science is everywhere – it is relevant to almost every unit on campus. This breadth provides both challenges and opportunities for IDEaS. To make progress, we support working groups and events in many focused areas and at the same time look for opportunities to avoid silos. We also leverage Georgia Tech’s vaunted position as leading technological university to drive forward big data driven science and engineering. We support data cyberinfrastructure needs of major centers such as the National Science Foundation (NSF) Cell Manufacturing Technologies (CMaT) Engineering Research Center and expect to serve upcoming projects in artificial intelligence (AI). Our services are valuable to many other IRIs and centers.
What couldn’t have happened without your IRI?
Catalyzation of Georgia Tech community in data engineering and science. The collective visibility is instrumental in winning large center-scale grants such as the South Big Data Regional Innovation Hub, the NSF TRIPODS institute, and the Hive supercomputer project. The IRI also exerts thought leadership at national and international levels and ensures GT has a seat at the table in U.S. data science leadership summits. For example, we organized the first US-Japan Big Data meeting for NSF.
What impact is your IRI GT research having on the world?
IDEaS has over 200 affiliated faculty working in many theoretical and applied areas. It is hard to summarize their impact in a just few words. Considering COVID situation as an example, we worked collaboratively with other big data regional hubs to launch COVID Information Commons for NSF, enabled faculty COVID research on HIVE which led to disease spread simulations advising the state, and supported events such as pandemic prevention workshop. I would also single out our management of the South Hub for outsized external impact. With 290 participating organizations, the Hub leads in community mobilization, tackling regional big data challenges, and education and workforce training.
]]>Utilizing Cyclodextrin to Compatibilize the Polymer and CNC Interface for Lightweight Material
Advisors: Will Gutekuhst; Professor, School of Chemistry and Mohan Srinivasarao; Professor, School of Materials Science & Engineering
As a renewable resource, cyclodextrin-modified CNCs have the potential to produce a new class of lightweight, high-strength composites for a wide-range of applications, including materials in the automotive and aerospace fields. The research this award will support involves using my recently developed surface modification technique to covalently anchor cyclodextrin rings onto cellulose nanocrystal (CNC) particles and study the physical threading of polymers through the cyclodextrin cavity
Receiving the IMat Graduate Fellowship is incredibly exciting and a great motivator as I am working on a challenging section of my project while working with lab restrictions. I am grateful to BASF for these funds, and I am excited to share my progress later in the year.
- Krista Bullard
Krista received her B.S. in chemistry at the University of Pittsburgh. While at Pitt, she conducted computational research on silyl ketene polymerization and CO2 absorption in ionic liquids with Dr. Daniel Lambrecht. During the summer of 2016, Krista received the Mickey Leland Energy Fellowship through the Department of Energy at the National Energy Technology Lab in Pittsburgh, where she did computational research of the electrochemistry of CO2 with gold nanoparticles. In the summer of 2017, Krista worked in polymer R&D at Sherwin-Williams. She is currently a PhD candidate in the School of Chemistry and Biochemistry and a recipient of the Renewable Bioproducts Institute fellowship.
Mechanocatalytic Ammonia Synthesis over Transition Metal Nitrides in Transient Microenvironments
Advisor: Carsten Sievers; Associate Professor, School of Chemical & Biomolecular Engineering
Hebisch and team aim to provide an alternative to the current way to produce fertilizers based on ammonia, which is currently highly industrialized and only viable at large scale, to enable increasing agricultural yields in developing regions. Their mechanocatalytic approach offers a promising alternative to industry use standards, as it can operate with renewable energy sources and features a simple, modular design.
The award of this fellowship shows that industry leaders also see a potential for our research to play a role in a more sustainable future and the funding provided will help to continue our pioneering work on this important topic.
- Karoline Hebisch
Karoline L. Hebisch is a second-year Ph.D. student in the School of Chemical & Biomolecular Engineering. She is advised by Dr. Carsten Sievers in the field of heterogeneous catalysis.
Karoline received her B.S. and M.S. in Chemistry from the Technical University of Darmstadt in 2016 and 2019. During her master’s she spent a semester abroad to study Plastics Engineering at the University of Massachusetts, Lowell. In her master’s thesis with Dr. Sievers, she studied mechanochemical ammonia synthesis as a sustainable, distributed approach for fertilizer production.
Developing and Understanding Liquid Metal Interfaces for Solid-State Batteries
Advisor: Matthew McDowell; Associate Professor, School of Mechanical Engineering and the School of Materials Science & Engineering
To make renewable energy possible for our society, progress needs to be made to improve energy storage devices. Fundamentally understanding the effects of liquid metals at solid- solid electrochemical interfaces will be an important step toward making solid-state batteries a competitive energy storage option. Our team aims to produce research results that help drive the development of batteries with higher energy density and specific energy that can be produced at scale for wide adoption.
Due to recent environmental disasters such as fires and hurricanes, climate change is again at the forefront of public discussion. New battery technologies will be critical for enabling longer- range electric vehicles and for the engineering of largescale energy storage technologies to mitigate greenhouse gas emissions for our society.
- Emily Klein
Emily Klein is a first-year graduate student in the School of Materials Science and Engineering, working in Prof. Matthew McDowell’s group and leading a research project focused on interfacial engineering of solid-state batteries. She has extensive prior research experience on investigating battery safety during her co-op experience at the Naval Research Lab, and is excited to be working to enable the creation of next generation solid-state batteries.
High-Throughput Screening of Cathode-Electrolyte Systems for Stable Lithium-Air Battery (LAB) Design using Machine-Learning (ML) and Density Functional Theory (DFT) Simulations
Advisor: Seung Soon Jang; Professor, School of Materials Science and Engineering
Machine Learning advances are aiding the development of reliable computational screening models for new energy storage devices. Unlike (de)intercalation in Lithium-ion batteries, Lithium-Air batteries requires an understanding of various competitive reactions taking place at both the cathode surface and solvated electrolyte. Training a machine learning model using data from molecular simulations can help us to predict battery performance for a wide array of substrates.
US Department of Energy aims to reduce price of batteries to about $80-100/kwh. There is an uprising need to develop cheap, energy dense, and compact devices. I am fortunate to study on fundamental mechanism of Lithium-air batteries, at this right hour. I am grateful to my advisor, Institute for Materials, and BASF for believing in and funding this project.
- Sai Saravanan Ambi Venkataramanan
Sai Saravanan Ambi Venkataramanan received his B.S in Chemical Engineering, with an emphasis on ASPEN and molecular simulations in studying the extraction efficiency of ionic liquids. He received the Indian Science Academies Summer Research Fellowship in 2018 and attended the CCP5-CECAM Summer Program in ‘Molecular Simulations of Condensed Phases’ at Lancaster University, UK. Sai is currently pursuing his M.S. in Materials Science and Engineering at Georgia Tech working under Prof. Seung Soon Jang in developing principal cathode and electrolyte specific features to predict Lithium-Air Battery performance.
The awardees will present their research to BASF representatives at BASF’s campus recruiting visit during October 2021. Fellows’ presentations will detail the aspects explored, new capabilities developed, and how their research impacts them personally and professionally, including benefits to their group, academic unit, Georgia Tech Community, and the larger society.
Founded in 2012 as one of Georgia Tech’s 11 interdisciplinary research institutes, the Institute for Materials at Georgia Tech seeks to enable and support Georgia Tech’s internationally recognized materials research and innovation ecosystem; establishing and supporting large- scale industry and government partnerships, developing opportunities for Georgia Tech researchers to catalyze new ideas, and establishing Georgia Tech as an internationally recognized hub for core materials research facilities, infrastructure and knowledge. Learn more at: https://research.gatech.edu/materials
The Georgia Institute of Technology, or Georgia Tech, is a top 10 public research university developing leaders who advance technology and improve the human condition.The Institute offers business, computing, design, engineering, liberal arts, and sciences degrees. Its nearly 40,000 students, representing 50 states and 149 countries, study at the main campus in Atlanta, at campuses in France and China, and through distance and online learning. As a leading technological university, Georgia Tech is an engine of economic development for Georgia, the Southeast, and the nation, conducting more than $1 billion dollars in research annually for government, industry, and society. For more on Georgia Tech research visit: https://research.gatech.edu/
BASF Corporation, headquartered in Florham Park, New Jersey, is the North American affiliate of BASF SE, Ludwigshafen, Germany. BASF has approximately 17,000 employees in North America and had sales of $18.7 billion in 2020. For more information about BASF’s North American operations, visit basf.com.
BASF, we create chemistry for a sustainable future. We combine economic success with environmental protection and social responsibility. More than 110,000 employees in the Group contribute to the success of our customers in nearly all sectors and almost every country in the world. Our portfolio is organized into six segments: Chemicals, Materials, Industrial Solutions, Surface Technologies, Nutrition & Care and Agricultural Solutions.BASF generated sales of €59 billion in 2020. BASF shares are traded on the stock exchange in Frankfurt (BAS) and as American Depositary Receipts in the U.S. Further information at basf.com.
- Christa M. Ernst - Interdisciplinary Research Communications Program Manager
Materials | Nanotechnology | Robotics
Georgia Institute of Technology| christa.ernst@research.gatech.edu
Dodds is the Flint Professor of Mathematics and Statistics at the University of Vermont. His research focuses on system-level big data problems in many areas, including language and stories, sociotechnical systems, contagion, and ecology. He also directs UVM’s Complex Systems Center and co-directs UVM’s Computational Story Lab.
As part of his visit, Dodds will deliver the IDEaS Distinguished Lecture on “The Science of Stories: Measuring and Exploring the Ecology of Human Stories with Lexical Instruments.” The event will be held on Wednesday, November 6 in the Klaus Advanced Computing Building, Rooms 1116 E & W, from 3:00-4:00 p.m.
Additionally, he will give another talk as a QBioS Special Seminar on “Rank Turbulence Divergence: A Tunable Instrument for Comparing Complex Systems.” The event will be held on Monday, November 11 in the Cherry Emerson Building, Room 320, from 12:15-1:15 p.m.
Dobbs will have office space in the Coda and Cherry Emerson Buildings to help facilitate interactions with both the IDEaS and QBioS research communities. You may find him in either location on November 4-15.
]]>
The National Science Foundation (NSF) announced the award, which will support the acquisition and deployment of the new supercomputer to TACC at The University of Texas at Austin, today. Anticipated to begin operations in 2019, Frontera will be the fastest at any U.S. university, among the most powerful in the world, and will allow the nation’s academic researchers to make important discoveries in all fields of science, from astrophysics to zoology.
The team from Georgia Tech is led by Srinivas Aluru, co-executive director of the Institute for Data Engineering and Science, and also includes Rich Vuduc, Edmond Chow, and David Bader. All are professors in the School of Computational Science and Engineering, where Bader serves as chair, and in the College of Computing.
“Georgia Tech is excited to be part of the NSF leadership-class computing facility project, which will produce the design and operation of the leading academic supercomputer of our times, and guide its path to reaching exascale capabilities,” Aluru said.
His team will be working on the evaluation of possible architectures and technical design of the Phase Two system, and develop exascale-ready code for application areas including computational biology and computational chemistry. Georgia Tech faculty will also serve on the technical advisory committee for the project.
“Many of the frontiers of research today can only be advanced by computing, and Frontera will be an important tool to solve grand challenges that will improve our nation’s health, well-being, competitiveness, and security,” said Dan Stanzione, TACC executive director.
If completed today, Frontera would be the fifth most powerful system in the world, the third fastest in the U.S., and the largest at any university. For comparison, Frontera will be roughly twice as powerful as Stampede2 (currently the fastest university supercomputer), and 70 times larger than Ranger, which operated until 2013.
To match what Frontera will compute in just one second, a person would have to perform one calculation every second for roughly one billion years.
Through its involvement, Georgia Tech researchers will influence the design, operation, and science conducted on the leading NSF supercomputer. Tech researchers will also receive substantial access during the early operations phase.
Anticipated early projects on Frontera include analyses of particle collisions from the Large Hadron Collider, global climate modeling, improved hurricane forecasting, and multi-messenger astronomy.
The primary computing system will be provided by Dell EMC and powered by Intel processors. Data Direct Networks will contribute the primary storage system and Mellanox will provide the high-performance interconnect for the machine. GRC (Green Revolution Cooling), NVIDIA, and the cloud providers Amazon, Google, and Microsoft will also have roles in the project.
Faculty at the Institute for Computational Engineering and Sciences at UT Austin will lead the world-class science applications and technology team, with partners from the California Institute of Technology, Cornell University, Princeton University, Stanford University, the University of Chicago, the University of Utah, and the University of California, Davis.
Experienced technologists and operations partners from the sites above, as well as The Ohio State University, the Georgia Institute of Technology, and Texas A&M University will ensure the system runs effectively in all areas, including security, user engagement, and workforce development.
Frontera will enter production in the summer of 2019 and will operate for five years. In addition to serving as a resource for the nation’s scientists and engineers, the award will support efforts to test and demonstrate the feasibility of an even larger future leadership-class system – ten times faster than Frontera – to potentially be deployed as Phase 2 of the project.
“Keeping the U.S. at the forefront of advanced computing capabilities and providing researchers across the country access to those resources are key elements in maintaining our status as a global leader in research and education,” said NSF Director France Córdova.
“This award is an investment in the entire U.S. research ecosystem that will enable leap-ahead discoveries.”
]]>jsalazar@gatech.edu
]]>The organizers who founded and spearheaded the Machine Learning in Science and Engineering (MLSE) Conference last year recently held the second-annual event June 9–12 on the Georgia Tech campus. Chaired by Dana Randall, Co-Executive Director of the Institute for Data Engineering and Science (IDEaS) and the ADVANCE Professor of Computing at Georgia Tech, the event brought together some of the brightest minds in machine learning from a variety of academic disciplines and institutions.
Machine learning spans a multitude of topic areas, therefore the conference attracted more than 400 attendees, including students, professors, and members of industry.
“There is a need for a conference like MLSE that focuses on the science and engineering domains that are being transformed by machine learning and AI,” said Randall. “All of the standard ML conferences are dominated by computer scientists and statisticians, so the many other scientists and engineers using ML never get critical mass to share challenges, opportunities, and progress.”
Georgia Tech’s Executive Vice President for Research Chaouki T. Abdallah launched MLSE 2019 with welcoming remarks before the lunchtime plenary talk. The conference included peer-reviewed talks and poster presentations from nine disciplines in which machine learning is playing an increasingly central role in new advances of scientific research: biomedical engineering, chemical engineering, chemistry, electrical and computer engineering, industrial engineering and operations research, materials science and engineering, mechanical engineering, physics, and public policy.
Last year, when the first MLSE Conference was held at Carnegie Mellon University (CMU), Randall and 2018 MLSE Chair Newell Washburn discovered more than 60% of the attendees responding to the post-conference questionnaire said they attended multiple tracks and benefited from talks in other fields in addition to their own. These results confirmed for Washburn and Randall that the format of bringing these vastly different groups of researchers together created a formula for a unique and beneficial ML conference.
MLSE 2019 enjoyed as much success as the inaugural year. “Several people came up to me to say they were blown away by how strong Georgia Tech is in machine learning across so many different fields,” Randall said. This conference showcased that with more than 42 talks, two of three short courses, and a hackathon, as well as several tracks all given or run by GT faculty and students, “Georgia Tech had large representation across every discipline represented in the nine parallel tracks.”
In addition to Georgia Tech and Carnegie Melon University, which had the most prominent showings, attendees hailed from more than 40 institutions of higher learning and delivered talks and presented posters over the three-day conference. Among the participants were faculty and students from Columbia University, Brown University, Duke University, Stanford University, Rutgers University, the University of California, Berkeley, and dozens of other preeminent institutions.
MLSE 2019 featured plenary talks by well-known experts in machine learning, including Jennifer Neville, the Miller Family Chair and associate professor of Computer Science and Statistics at Purdue University; Eliu Antonio Huerta Escudero, head of the Gravity Group at the National Center for Supercomputing Applications (NCSA) at the University of Illinois at Urbana-Champaign; John F. McDonald, professor in the School of Biology at Georgia Tech; Daniel B. Neill, associate professor of computer science, public service, and urban analytics at New York University; and Ross Thomson, solutions architect for Scientific Computing at Google Cloud Platform of Google. Their topics spanned relational artificial intelligence, astrophysics, cancer research, and policy.
In addition to the nine parallel tracks and two poster sessions, MLSE 2019 offered three short courses on Sunday, June 9, held in conjunction with a day-long Women in Data Science Workshop (WDSW). Two of these courses were led by Georgia Tech professors: Yao Xie, the Harold R. and Mary Anne Nash Early Career Professor and assistant professor in the Stewart School of Industrial and Systems Engineering and Zsolt Kira, associate director of the Machine Learning Center at Georgia Tech, assistant professor in the School of Interactive Computing, and research scientist at the Georgia Tech Research Institute.
The WDSW was attended by almost 200 women from across the southeast and beyond. This workshop—together with MLSE 2019—formed the Annual Data Science Forum. Many of the women who attended the workshop remained in Atlanta to attend the MLSE conference, helping to make MLSE a more inclusive and diverse conference. Travel grants were provided to more than 60 students to attend one or both events composing the forum.
The WDSW featured a keynote talk by Lo Li, chief technology officer at Equifax, and talks by other prominent women in data science, such as Ellen Zegura of Georgia Tech, Jennifer Priestley of Kennesaw State University, Lakshmi V. Kalluri of Anthem, and Brandeis Marshall of Spelman College. Discussions following some of the talks gave attendees a platform for considering the opportunities for women in the field. The workshop concluded with the three parallel short courses, also attended by MLSE participants.
This year’s Annual Data Science Forum received generous support from NSF through grant number 1839340 and from the Microsoft Corporation, along with additional funding from several other companies.
The third MLSE event will be held in New York City, hosted by Columbia University. Details about next year’s conference will be posted on the IDEaS website as they become available.
]]>“I didn’t want to go anywhere else,” he said. He had grown up in Atlanta, graduating from Mays High School, and he loved the city. More than that, he already knew that he wanted to work with computers, and he knew Georgia Tech was one of the best places in the world to do so.
When he got to campus, he knew right away that he had made a good decision.
“I always felt I belonged at Georgia Tech,” Isbell said. “No, I didn’t join a frat, I wasn’t part of any of the big clubs,” he said. “Hey, I went to zero parties. Zero. But I did build friendships. I built connections.” He also, in a nice bit of symmetry, served as the undergraduate representative on the committee that hired Peter Freeman to be the first dean of the brand new College of Computing.
Today, Charles Isbell becomes the John P. Imlay Jr. Dean of Computing. He is the fourth person to hold the position. His philosophy as dean is built on the foundation he laid long ago as an undergraduate.
“To me, it’s all about community,” he said. “I want people to feel like they belong, and that the community reflects their experiences. I want people to feel that the things they’re learning apply to their worlds.”
Machines Bringing People Together
Isbell went to MIT after graduating from Georgia Tech, and after that spent four years working at AT&T Labs. During that time, he continued to pursue his interests in computing and human connection.
The first project that earned Isbell a “best paper” award was his work on Cobot, a software agent whose goal was to become a functioning member of an online social community called LambdaMOO.
“I’m interested in how humans express themselves in a way that computers can understand – from a technical, machine learning point of view, that is,” Isbell said.
He also found new ways to use technology to serve existing real-life communities. At MIT, he built what was most likely the first-ever online Black history database. He ran a website for hip-hop reviews.
To this day, he continues to mix his cultural experience and computing. All of his graduating students pose for photos dressed like members of the funk band Parliament in a silver top-hat, star-shaped sunglasses, and strings of Mardi Gras beads. The framed and funky photos line the walls of his office.
Isbell says that combining his passions keeps him engaged and that he likes to see others do the same.
“If you’re passionate, people pick up on that passion,” he said.
“The technology we develop is transformative, and we have to reckon with that. We have to accept our responsibility as leaders and our responsibility to bring other people along for this ride.”
Charles Isbell, John P. Imlay Jr. Dean of Computing
Giving Back
In 2002, Isbell was hired as a junior faculty member in the College of Computing and moved back to Atlanta.
“First thing that happened when I came back, my mother made me a bowl of cheese grits and bacon,” he said. “I knew I was back home.”
At the Institute, things were more complicated.
“It was so exciting to be back, but the place was completely different,” he said. It was bigger, a stronger program with a ballooning reputation. “Still, I always felt I could build something here.”
After earning tenure, Isbell dived into administrative work to do exactly that. He was one of the architects of the college’s award-winning Threads curriculum, and also of its groundbreaking Online Master’s of Science in Computer Science (OMSCS) program.
“I just kept volunteering,” he said. “Then one day I woke up as dean.”
Redefining the Field
He didn’t, of course. Wake up as dean, that is. Isbell won the job in a grueling nationwide search. He is the first internal candidate ever to be named as the dean of the College of Computing.
And as someone who has been in or around the college for decades, he has a unique view on its development. When Isbell arrived as an undergraduate, computing was still in its infancy at Georgia Tech — it wasn’t even a college yet.
Through his tenure on the faculty, he has seen the college grow and mature. Now, he says, the college is truly entering adulthood, a top-10 program with responsibilities not only to its faculty, staff, and students but also to the larger world.
“The technology we develop is transformative, and we have to reckon with that,” he said. “We have to accept our responsibility as leaders and our responsibility to bring other people along for this ride.”
As dean, he intends to build on the hard work of his predecessors in confronting the challenges of a field that is always changing and always short of labor. And as computing metastasizes into other fields – finance, health, media, politics, art — he sees social and ethical considerations becoming ever more important.
The good news is that the College of Computing is already addressing these problems, Isbell said. OMSCS has diversified and significantly increased the pipeline of trained talent to industry. Our Constellations Center for Equity in Computing is piloting a hybrid classroom-online model that holds the promise of making computer science education available to all children. The college has made ground-breaking commitments to not only teach ethics to the students, but to computing research that prioritizes transparency and the public good.
In other words, Isbell wants Georgia Tech to lead a re-thinking of the nature and importance of community in the field of computing.
“It seems increasingly clear that computer scientists need to think more clearly about the impact of their work on society as a whole,” Isbell said. “That’s going to require the involvement of everyone who is affected — which is to say, everyone.”
]]>The funding will continue support for the South Big Data Innovation Hub, an organization that helps 16 Southern States and the District of Columbia identify and utilize data science and engineering to address critical societal needs. One of four NSF-supported regional data hubs in the U.S., the South Big Data Hub is managed by the Georgia Institute of Technology and the University of North Carolina-Chapel Hill.
"The Big Data Hubs provide a connective tissue for the data science ecosystem across sectors and domains,” said Renata Rawlings-Goss, the Hub’s executive director. “I am deeply pleased by NSF's recommitment to the growth of the South Hub and our community. Over the last three years, we have made great strides within our priority areas and are looking to broaden that reach in the next four years.”
The NSF-supported data hubs play four key roles: (1) Accelerating public-private partnerships that break down barriers between industry, academia and government, (2) Growing R&D communities that connect data scientists with domain scientists and practitioners, (3) Facilitating data sharing and shared cyber infrastructure and services, and (4) Building data science capacity for education and workforce development.
“There is a global shortage of data science and analytics talent that is threatening the future of innovation,” added Rawlings-Goss “By working across sectors, the South Hub joins in creating solutions to increase the capacity of universities and industry to work on pressing problems for our region and for the world.”
Priorities for the hubs are determined regionally to bring together collaborators that include academics, community leaders, local and state government executives, regional businesses, national laboratories and others, explained Srinivas Aluru, principal investigator for the Hub, which was launched in 2015 and won the 2019 Georgia Tech Outstanding Achievement in Research Development Award.
“We want to collaborate to help solve regional problems using the resources of the Hub,” explained Aluru, who is also co-executive director of the Institute for Data Engineering and Science at Georgia Tech. “We are addressing truly regional issues that affect more than one state and more than one set of collaborators. These are challenges that can only be addressed by bringing these groups together.”
The south region is pursuing five major big data priorities:
The new NSF award includes seed funding designed to evaluate the feasibility of new big data projects. Part of a hub-and-spoke system, the seed money should help create new spokes to address specific data issues identified by collaborators.
“Developing innovative, effective solutions to grand challenges requires linking scientists and engineers with local communities,” said Jim Kurose, Assistant Director for Computer and Information Science and Engineering at the NSF. “The Big Data Hubs provide the glue to achieve those links, bringing together teams of data science researchers with cities, municipalities and anchor institutions.”
Ultimately, the goal is to harness the synergy of the collaborators to address issues that require the use of data science and engineering techniques.
“By catalyzing partnerships that integrate academic researchers into the fabric of communities across the U.S., we can accelerate and deepen the impact of basic research on a range of societal issues, from water management to efficient transportation systems,” said Beth Plale, one of the NSF program directors managing the Big Data Hubs awards.
The South Big Data Hub was funded through the National Science Foundation’s Big Data Science & Engineering Program, Awards 1550305 and 1550291. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.
Research News
Georgia Institute of Technology
177 North Avenue
Atlanta, Georgia 30332-0181 USA
Media Relations Contact: John Toon (404-894-6986) (jtoon@gatech.edu).
Writer: John Toon
]]>Research News
(404) 894-6986
]]>“We were proud to honor the leadership of the South Big Data Regional Hub this year,” said Chaouki T. Abdallah, Georgia Tech’s executive vice president for Research. “The SBDH provides national visibility for Georgia Tech as a thought leader in big data and further positions the Institute as a regional center of excellence.”
The Georgia Tech team members recognized for their leadership in this program include:
Srinivas Aluru, professor in the School of Computational Science and Engineering (CSE) and co-executive director of the Institute for Data Engineering and Science (IDEaS);
Renata Rawlings-Goss, executive director of the South Big Data Hub;
Ashok Goel, professor in the School of Interactive Computing; and
Gari Clifford, adjunct professor in the Wallace H. Coulter Department of Biomedical Engineering.
"It is extremely gratifying to launch the South Big Data Hub and nurture it to grow into a national asset,” said Aluru, primary investigator of SBDH. “The Hub community is broad, undertaking regionally and nationally important challenges in diverse areas including health disparities, environmental sustainability, energy infrastructure, coastal hazard management, and data science education and workforce training.”
The SBDH was launched in 2015 as part of the National Science Foundation’s Big Data Research and Development Initiative. The SBDH is jointly coordinated with the University of North Carolina and includes broad membership from academia, industry, government, and nonprofit organizations across 16 southern states and Washington, D.C.
The award was presented to the SBDH leadership team at the Institute Faculty and Staff Honors Luncheon last week, shortly after the South Big Data Regional Innovation Hub hosted the annual All-hands Meeting.
[Related Links: Faculty, Staff Honored at 2019 Luncheon]
]]>Georgia Tech’s Executive Vice President for Research Chaouki T. Abdallah launched the meeting with welcoming remarks, including stressing the importance of the SBDH for “taking scientific approaches to solving social issues, including those in healthcare and smart cities.” He lauded the hub’s work in “outstanding research on smart grids, a smartphone app for healthcare, and the DataUp program.”
The accomplishments Abdallah referenced hail from research coming from SBDH spoke projects, which are each funded with a $1 million award to focus on regional challenges. The Smart Grids Big Data group project includes 52 academic partners, 32 industry service providers and utilities, and 22 government and non-profit partners. The group discovered the potential for major improvements in smart grid operation.
Another spoke project, Large Scale Medical Informatics for Patient Care Coordination and Engagement, includes a six-university team that developed a smartphone app and cloud infrastructure to analyze patient behavior, environment, and health to provide insights into health effects of daily behaviors, with the goal of understanding and mitigating health disparities in African American and Hispanic populations.
And finally, SBDH Executive Director Renata Rawlings-Goss led the DataUp program to increase the capacity of Southern schools to teach data science, particularly at minority-serving institutions, community colleges, and four-year liberal arts colleges. The effort resulted in a 50-page report that provided integral data and information for the National Academies of Sciences, Engineering, and Medicine's project Envisioning the Data Science Discipline: The Undergraduate Perspective.
Attended by more than 100 individuals, the meeting included participation by researchers from key universities in the Southern region, representatives from local companies in the healthcare and consulting industries, as well as members of government entities.
“The SBDH serves 16 Southern states from Delaware to Texas, and Washington D.C. with more than 1,000 members from 290 organizations, including academia, industry, government, and non-profits,” said Srinivas Aluru, the principal investigator (PI) for the SBDH, co-executive director of the Institute for Data Engineering and Science, and a professor in the School of Computational Science and Engineering. “The all-hands meeting is one of the most important events we hold as it is integral to fostering R&D collaborations in the thematic areas of interest to the SBDH region.”
An important component of the first day of the meeting was a panel discussion by the PIs for the six spoke visionary projects. The spoke leaders include three Georgia Tech faculty members: Ashok K. Goel from the School of Interactive Computing, Gari Clifford from the Department of Biomedical Engineering and interim chair of Biomedical Informatics at Emory University, and Santiago Grijalva, the Georgia Power Distinguished Professor of Electrical and Computer Engineering.
Day one also featured a panel on priority areas composed of the hub’s co-PIs and priority area representatives that focused on an effort to launch new initiatives in the areas of greatest need. Breakout sessions, a student poster session, and a keynote speaker at dinner rounded out the first day of the all-hands meeting.
Day two of the all-hands meeting included working groups on cyberinfrastructure and social cybersecurity, as well as an industry panel with representatives from Anthem, Microsoft Research, Stanley Black & Decker, and others.
The successful event concluded with meetings for steering committee members, who serve as an advisory group for the hub and help disburse seed funding, as well as meetings amongst member representatives from the other three regional hubs funded by the National Science Foundation.
]]>Galil’s deanship was marked by accomplishments on many fronts. Under his leadership the College has risen into the top eight nationally, top seven internationally – the only top 10 computer science program to rise either in rank or in score in the last ranking (2018).
In a measure of the College’s public perception, applications to the College have grown ten-fold, and enrollment in on-campus degree programs has nearly doubled during Galil’s tenure as dean. Computing is now the largest major at the university, and the most selective – our majors average higher than 1500 on the SATs.
[RELATED: College of Computing Rises to No. 8 in U.S. News Rankings]
The College’s reputation among employers and alumni has seen dramatic enhancement, as well. As a result, the College's career fairs and its corporate affiliates program have grown in stature in recent years. The GT Computing Career Fair regularly sets new attendance records with more than 160 companies participating (with nearly 20 companies waitlisted) this year in the Klaus Building Atrium. Several hundred students from across campus attended each day of the four-day event.
“Being a dean is about community building, about involvement, support, and empowerment. You’re closer to students, you’re closer to staff and faculty. I view my role as dean as working to inspire our community by helping them to connect, encouraging them to excel, increasing their confidence.” - Zvi Galil
More and more companies are also participating in the College's corporate affiliates program (CAP). During Galil's tenure as dean, CAP grew from 14 companies generating $280,000 in membership fees in 2010, to 63 companies raising $1.13 million in the current academic year. Galil exceeded the annual campus fundraising campaign goal by 40 percent – the largest percentage above the goal of any unit at Georgia Tech. Alumnus James Liang's gift of $1.5 million for an endowed chair was at the time the largest international gift in Georgia Tech history, and the only endowed chair by an international donor.
The successful fundraising provided the resources for continued investment in the College and its faculty, and also helped fund four Interdisciplinary Research Institutes and four Interdisciplinary Research Centers led by the College. Galil doubled the number of endowed senior faculty chairs to 10, in addition to four new junior faculty chairs. Faculty rose from 85 to 102, with six or more to join later this year.
[RELATED: Corporate Affiliates Program Paying Off for GT Computing Students]
It is a testament to the values and productivity of the College’s faculty that, with just 8 percent of Georgia Tech faculty, GT Computing teaches about 18 percent of the Institute’s credit hours (about 13 percent of undergraduate and about 24 percent of graduate credit hours).
“Under Zvi’s leadership the standing of the college has improved along a host of traditional metrics – but truly great universities are in the center of the important issues of the day,” said Executive Associate Dean Charles Isbell, who will take over as dean on July 1. “Through OMSCS, Zvi has led the way in moving the college to the center of perhaps the most important of national discussions: the role of affordability and access in computing. That is a transformative accomplishment.”
OMSCS
Creating the College’s now-famous Online Masters of Science in Computer Science (OMSCS) program took years of labor from dozens of faculty and staff members. Galil’s vision was the driving force behind the entire project, however, and guided many of the decisions that make the program so distinctive.
One of the most significant was OMSCS’ unique admissions policy. Instead of admitting only a few of the highest-achieving applicants, Galil insisted that the program be open to anyone who had met the requirements. Those online students have been just as successful as the on-campus students admitted through a much more selective process.
Now, five years after its founding, the online master’s has nearly 9,000 students and an international reputation for changing the game in online education. The OMSCS program embodies Georgia Tech's motto of Progress and Service with its unique combination of prestige, accessibility, and affordability. Its launch has changed national and international perspectives on Georgia Tech.
[RELATED: Juggling Careers, Grad School, Kids: One Family’s Story of How They Make OMSCS Work]
“OMSCS offers wider access to the high quality of our residential program at a substantially lower cost. It helps realign today’s workforce with the requirements of a thriving 21st-century economy. This is a fundamental, revolutionary shift from the prevailing paradigm of higher education, in which a brand is bolstered by exclusion and high tuition fees,” Galil said.
Building a community
“Being a dean is about community building, about involvement, support, and empowerment,” Galil said. “You’re closer to students, you’re closer to staff and faculty. I view my role as dean as working to inspire our community by helping them to connect, encouraging them to excel, increasing their confidence.”
Galil has made particular efforts to integrate staff members into the community – through regular meetings and an annual staff retreat – and is well known for matching high standards with a collaborative approach and approachability.
“Zvi pushes for excellence in a way that stretches everyone,” said Alan Katz, assistant dean for finances and administration. “He believes in sharing information, serving others, and providing incentives – he’s a carrot person, not a stick person.”
“You would never know Zvi has such a high status because he’s so down to earth,” said Pam Ruffin, director of human resources for the college. “You can walk up to his door and he’ll take time to talk to you.”
Even those who don’t make it to his office hear from Galil regularly, through a steady stream of e-mails he sends out to the entire GT Computing community. Although he is known as “the e-mail dean,” he almost never mentions himself in his missives. “I love to brag about the achievements of faculty, staff, and students,” he said. “I want everyone to know they are the most important part of the College.”
In parting
Galil – a highly influential scholar in the fields of algorithmic design and analysis, and computational complexity and cryptography – is a member of the National Academy of Engineering, and a fellow of the Association for Computing Machinery and of the American Academy of Arts & Sciences. Prior to coming to Georgia Tech, he served as the dean of engineering at Columbia University and the president of Tel Aviv University.
Yet, he views his deanship at GT Computing as the most satisfying period of his career.
“In OMSCS, we pioneered a program that proved high-quality, cost-reduced online education at scale is doable, and that it satisfies an unmet need – being radically more accessible and affordable than on-campus options,” Galil said. “I view it as my greatest achievement.”
As for his message to GT Computing faculty, staff, students, and alumni, “GO JACKETS!”
]]>DoD’s Multidisciplinary University Research Initiatives (MURI) Program funds projects that bring researchers together from diverse backgrounds to work on a complex problem. Institute for Data Engineering and Science co-director, Professor Dana Randall, is project investigator and leads a team of six that includes Daniel Goldman, Dunn Family Professor in the School of Physics. The Formal Foundations of Algorithmic Matter and Emergent Computation team also includes chemical engineering, mechanical engineering, physics, and computational science researchers from other universities.
The researchers are trying to predict and design emergent behavior within computation by using basic algorithms on simple machines to perform complex tasks. Emergent behavior is when a microscopic change in a parameter creates a macroscopic change to a system. This collective behavior is easy to find in nature, from a swarm of bees to a colony of ants, but also appears in other scientific disciplines.
“A MURI lets us take a deep dive toward understanding how many computationally limited components at the micro-scale can be programmed to work collectively to produce useful behavior at the macro-scale,” said Randall, who is also the ADVANCE Professor of Computing. “Our interdisciplinary team combines expertise in many fields, mimicking the research by forming a collaboration that is also greater than the sum of its parts."
The MURI hybrid approach to algorithmic matter combines traditional logic-based programming with non-traditional computational methods, such as using physical characteristics of the interacting matter to drive a system toward collective behavior. One of the goals is to program based on this predictable emergent behavior. The approach also predicts basic properties of the collective’s emergent behavior, like whether it will behave like a gas, fluid, or solid. In this context, emergent behavior turns into emergent collective computation.
“MURI promises basic algorithms that allow very simple machines to work collectively to perform amazingly complex tasks,” Massachusetts Institute of Technology (MIT) chemical engineering Professor Michael Strano said. “Our team will examine systems of autonomous cell-like particles that interact and respond to the movement of their neighbors in a programmable way. Theorists will be able to test ideas of emergent computation from these simple devices and learn how to execute tasks from the behavior of relatively simple, autonomous particles.”
Although the behavior has footing in physics, computer science, and swarm robotics, there is no underlying framework to explain why until this research. The multidisciplinary approach allows theory and experiment to continuously inform each other and determine the computational capabilities of emergent behavior. The team has an ideal range of expertise in machine learning, control theory, and non-equilibrium physics and algorithms. They are also working with experimentalists who build collective systems at granular and microscopic scales.
“An exciting aspect of this collaboration will be our attempts to interface and integrate ideas and tools from robotics, non-equilibrium physics, control theory, and computer science to develop task-capable swarms,” Goldman said.
This MURI project will run for five years and is funded by the Army Research Office. In addition to Randall, Goldman, and Strano, the team also includes Arizona State computational science and engineering Professor Andrea Richa, MIT physics Associate Professor Jeremy England, and Northwestern mechanical engineering Professor Todd Murphey.
The overarching goal is to find how simplistic the computation can be for this complexity. This could lead to advances in engineered systems achieving specific task-oriented goals.
“The MURI promises nothing short of the transformation of robots,” Strano said, “from the large, bulky constructions that we think of today, to future clouds or swarms that enable functions that are currently impossible to realize.”
Writer: Tess Malone
]]>College of Computing
]]>As the volume of data continues to expand, however, many agencies are considering the use of commercial cloud services to help store and make it available to users. While agencies may have different strategies, these new partnerships could result in user fees levied on downloads and analyses performed on the data while it remains in the cloud.
Writing in a policy forum article published February 8 in the journal Science, a Georgia Institute of Technology space policy researcher who studies such data use urges caution about the design of these commercial cloud partnerships and possible imposition of user fees.
“Under the current system, free and open government data is used by scientists to conduct research, by entrepreneurs to create new businesses, and by citizens and other organizations to promote government transparency,” said Mariel Borowitz, an assistant professor in Georgia Tech’s Sam Nunn School of International Affairs. “If users must pay fees to download or analyze the data, this will decrease the ability of these users to access and work with data. Past experience suggest that the impacts of this decrease in data use could be large – both for individual users and for society as a whole.”
Moving data to commercial cloud systems would likely provide broader access and more efficient analysis options, but she cautions those advantages could be offset by the cost, particularly for organizations with small budgets.
“Agencies risk losing some of the benefits of this transition by not budgeting for the costs associated with data downloads and analysis, up to a reasonable level,” Borowitz said. “Many who would be interested in using the data may not be able to pay the associated fees. Researchers, nonprofit organizations and others who do not directly profit from the use of this data are most likely to be affected.”
Borowitz recently spent two years at NASA and witnessed both the development of systems that will dramatically increase data collection and debates about future data storage. She recently authored a book, Open Space: The Global Effort for Open Access to Environmental Satellite Data, published by MIT Press.
She would like to see the agencies that provide data continue to shoulder the costs, up to some “reasonable level,” to ensure that the data continues to be readily available to all users. As an alternative to commercial services, some agencies are considering development of their own, custom-built cloud solutions, and will have to weigh the cost of benefits of the different options. There will also be technical, organizational and policy issues to consider.
“Agencies are taking seriously issues of security and long-term preservation of data,” Borowitz added. “When working with commercial providers, some are concerned about the possibility of getting ‘locked in’ to one provider, due to the large costs of migrating data from one system to another. It is possible that costs and capabilities could change over time. On the other hand, commercial cloud providers have large workforces and extensive infrastructure that allow them to provide services and capabilities well beyond what any one agency would be able to maintain.”
Borowitz notes that most agencies have not made final decisions about their cloud-based programs, so there should be adequate time to work through these issues.
“Most agencies that make data publicly available, particularly science agencies, are already discussing and/or beginning to make the transition to cloud systems,” she said. “However, these programs – at agencies like NSF, NIH, NASA and NOAA – are still in their early phases, and there is still opportunity for feedback to be provided and adjustments to the programs to be made.”
The existence of fees for access to government data is not without precedent, but Borowitz argues that past experience suggests that user fees result in significantly less use. Before Landsat data – satellite imagery of Earth – was made freely available in 2008, no more than 25,000 images a year were purchased from the collection. “Within a few years of implementing the free and open data policy, the government was distributing 250,000 images a month,” she said.
That number provides a suggestion of what the often cash-strapped agencies are dealing with. According to the paper, the National Oceanic and Atmospheric Administration (NOAA) houses more than 100 petabytes (PB) of data and generates more than 30 PB per year from satellites, radars, computer models and other sources. NASA projects that its archive will grow to 250 PB by 2025. And the amount of genomic data at the National Institutes of Health is growing exponentially.
A petabyte is 1,024 terabytes, or a million gigabytes. A gigabyte is 1,024 megabtyes. For scale, an average photograph taken by a high-end cell phone camera can be in the neighborhood of 10 megabytes. Laptop computers may be able to store as much as a few terabytes of data.
Borowitz sees the transition to cloud computing as both an opportunity and a challenge for the future availability of government data. “The decisions being made right now about the structure of these programs have the potential to significantly impact researchers and society as a whole, so it is important to raise awareness and increase engagement on these issues.”
CITATION: Mariel Borowitz, “Government data, commercial cloud: Will public access suffer?” (Science, 2019) http://science.sciencemag.org/content/363/6427/588
Research News
Georgia Institute of Technology
177 North Avenue
Atlanta, Georgia 30332-0181 USA
Media Relations Contact: John Toon (404-894-6986) (jtoon@gatech.edu).
Writer: John Toon
]]>]]>
Research News
(404) 894-6986
]]>In response to an email query about AI developments that can be expected in 2019, most of the researchers – whether talking about machine learning (ML), robotics, data visualizations, natural language processing, or other facets of AI – touched on the growing importance of recognizing the needs of people in AI systems.
“In 2019, I hope we will see AI researchers and practitioners start to frame the debate about proper and improper uses of artificial intelligence and machine learning in terms of human rights,” said Associate Professor Mark Riedl.
[RELATED: Is AI Coming For My Job?]
“More and more, interpretability and fairness are being recognized as critical issues to address to ensure AI appropriately interacts with society,” said Ph.D. student Fred Hohman.
Questions about the rights of end users of AI-enabled services and products are becoming a priority, but Riedl said more is needed.
“Companies are making progress in recognizing that AI systems may be biased in prejudicial ways. [However,] we need to start talking about the next step: remedy. How do people seek remedy if they believe an AI system made a wrong decision?” said Riedl.
Assistant Professor Jamie Morgenstern sees algorithmic bias as an ongoing concern in 2019 and gave banking as an example of an industry that may be in the news for its algorithmic decision-making.
“I project that we’ll have more high-profile examples of financial systems that use machine learning having worse rates of lending to women, people of color, and other communities historically underrepresented in the ‘standard’ American economic system,” Morgenstern said.
[RELATED: Researchers Working To Improve Fairness in the ML Pipeline]
In recent years corporate responses to cases of bias have been hit or miss, but Assistant Professor Munmun De Choudhury said 2019 may see a shift in how tech companies balance their shareholders’ interests with the interests of their customers and society.
“[Companies] will be increasingly subject to governmental regulation and will be forced to come up with safeguards to address misuse and abuse of their technologies, and will even consider broader partnerships with their market competitors to achieve this. For some corporations, business interests may take a backseat to ethics until they regain customer trust,” said De Choudhury.
One way companies can regain that trust is through sharing their algorithms with the public, our experts said.
“Developers tend to walk around feeling objective because ‘it’s the algorithm that is determining the answer’. Moving forward, I believe that the algorithms will have to be increasingly ‘inspectable’ and developers will have to explain their answers,” Executive Associate Dean and Professor Charles Isbell.
Ph.D. student Yuval Pinter agreed. In the coming year, “[I] think we will see that researchers are trying to [develop] techniques and tests that can help us to better understand what’s going on in the actual wiring of our very fancy machine learning models.
“This is not only for curiosity but also because legal applications or regulation in various countries are starting to require that algorithmic decision-making programs be able to explain why they are doing what they are doing,” said Pinter.
Regents’ Professor Ron Arkin believes that these concerns are becoming more central precisely because artificial intelligence will continue to grow in importance in our everyday lives.
[RELATED: Who's Behind the Wheel?]
“Despite continued hype and omnipresent doomsayers, panic and fear over the growth of AI and robotics should begin to subside in 2019 as the benefits to people’s lives are becoming more apparent to the world.
“However, I expect to see lawyers jumping into the fray so we may also see lawsuits determining policy for self-driving cars [and other applications] more so than government regulation or the legal system,” said Arkin.
]]>To help further this process of defining data science, dozens of leading data scientists recently gathered in Park City, Utah for the 2018 Data Science Leadership Summit.
The goal for the summit was to create an academic community for data science that takes collective responsibility for advancing research, education, practice, outreach, and public engagement. School of Computational Science and Engineering Professor and Co-Executive Director of the Institute for Data Engineering and Science Srinivas Aluru served as chair of this historic summit.
“Most universities don’t have data science departments. Essentially, we want to identify who is leading data science activities on these campuses. In some cases, it’s an institute, in some cases it’s a person leading an educational program, and in some cases, it could be a department chair,” said Aluru.
“We are looking for people who are influencing data science research and education and bringing them together for a collaborative conversation.”
The two-day summit, held Oct. 12-13, included session topics ranging from data science organizational structures, undergraduate data science education to data ethics and privacy as well as academic data science career paths. Two keynote presentations by leading data scientists Vipin Kumar, of the University of Minnesota, who recently presented a CSE distinguished lecture, and Blaise Aguera y Arcas of Google AI were also highlights of the event.
The summit marks the second recent gathering of academic leaders who influence data science research and education on their campuses.
International data science event
A similar meeting targeted to data science institute and center leaders worldwide took place in August. The day-long invitation-only applied data science panel was orchestrated by the Alan Turing Institute and The Data Science Institute of the Imperial College London. Aluru represented Georgia Tech at the event, which was held in conjunction with the twenty-fourth international conference on Knowledge Discovery and Data Mining (KDD 2018).
During the event, attendees explored challenges to understanding and defining standards and definitions for analytics roles, skill sets, and career paths in the data science industry.
“Some of the goals of the day were to learn best practices from other institutes and programs, figure out how to move forward together as a community, and understand the funding models across different countries and universities. We also made progress on issues such as how should we educate? What research we should focus on? And, how best can education programs, industry partnerships, and research work together for the overall benefit of the field,” said Aluru.
As part of the conversation and educational aspect of the panel, each attendee presented a poster detailing how their data science institute was run, how they were funded, and what the different roles and responsibilities were for their respective institutions.
“As the new field of data science expands, the need to identify and define both its use and direction becomes paramount to this seemingly large and omnipresent field,” said Moderator Usama Fayyad in the panel description.
]]>Communications Officer
]]>Deep learning has been a major driving force in the recent surge of interest in artificial intelligence (AI) and machine learning. While deep learning has resulted in tremendous empirical success, the theoretical understanding of deep learning remains an important research field rife with opportunity.
“In other words, understanding why and how AI works or doesn’t work is not clear,” said Xiaoming Huo, a professor in the H. Milton Stewart School of Industrial and Systems Engineering (ISyE) and executive director of TRIAD.
“The theory needs significant development, so the workshop brought together people who have been working on this problem and enabled them to exchange ideas,” he noted.
The workshop’s interdisciplinary focus was reflected in both the diversity of speakers, including experts in fields ranging from statistics and optimization to computer science and applied mathematics, and its sponsors, which included ISyE; Georgia Tech’s colleges of Computing, Engineering, and Sciences; Hong Kong University of Science and Technology’s Big Data Institute; the Statistical and Applied Mathematical Sciences Institute (SAMSI); and the American Statistical Association (ASA).
Promising ideas on the theoretical foundation of deep learning have emerged, and the TRIAD workshop provided an avenue for researchers in related fields to review existing work, communicate new results, and seek new research directions.
About TRIAD
TRIAD is a cross-disciplinary institute that was established in 2017 as part of the National Science Foundation’s TRIPODS (Transdisciplinary Research in Principles of Data Science) program. TRIAD unites statistics, mathematics, and theoretical computer science to further develop the foundations of data science. TRIAD brings together senior, mid-career, and junior faculty members, postdoctoral fellows, graduate and undergraduate students, all from Tech’s colleges of Computing, Engineering, and Sciences, and data science practitioners-at-large using focused working groups, national and international workshops, and organized innovation labs.
]]>H. Milton Stewart School of Industrial and Systems Engineering
404.385.4745
]]>The awards were designed to expand the scope of the TRIPODS cross-disciplinary institutes established last year, including Georgia Tech’s Transdisciplinary Research Institute for Advancing Data Science (TRIAD).
“The multidisciplinary approach for addressing the increasing volume and complexity of data enabled through the TRIPODS+X projects will have a profound impact on the field of data science and its use,” said Jim Kurose, NSF assistant director for Computer and Information Science and Engineering (CISE). "This impact will be sure to grow as data continues to drive scientific discovery and innovation.”
A total of $8.5 million in TRIPODS+X grants were awarded this year, supporting 19 collaborative projects at 23 universities, and bringing new perspectives to complex and entrenched data science problems in science, engineering, and mathematics.
The three Georgia Tech projects span three different NSF priorities in education, visualization, and research.
Education: Data-driven Discovery and Alliance
Prasad Tetali and his multi-institutional team including traditional women’s and historic black colleges and universities, are developing undergraduate courses for STEM majors to give more students access to a data-driven future. With this grant, the collaborative alliance, grounded in math, statistics and computer science theory, will develop a toolkit of data science modules to integrate into science curriculum at Agnes Scott College, Morehouse College, and Spelman College. They will also hold boot camps and workshops. The educational outreach will enrich the knowledge of these institutions’ faculty, and later, the team plans to adapt the initiative to serve other research-intensive women’s and HBCU institutions.
“The NSF-supported educational alliance is exciting in many ways,” Tetali says. “It gives opportunity to infuse the foundational data science curriculum with real-world applications from the physical and life sciences. It will also likely catalyze collaborative research in data science and related fields between Georgia Tech and Atlanta area colleges.”
Investigators:
Award Amount: $200,000
Vision: Creating an Annual Data Science Forum
Dana Randall and colleagues from Carnegie Mellon University and Columbia University are creating a week-long Data Science Forum built around the Second Symposium on Machine Learning in Science and Engineering (MLSE). The forum combines multiple events aimed at catalyzing communication across foundations, applications, and disciplinary fields, and at fostering diversity and inclusion. Two new workshops that complement the conference are a part of the forum: A Women in Data Science Workshop, and a Foundations of Data Driven Discovery workshop.
MLSE, begun last year by Georgia Tech and Carnegie Mellon, was the first annual machine learning conference organized to collocate tracks within traditional disciplines using machine learning while allowing an exchange of ideas across disciplines. This cross-disciplinary breadth combined with efforts to build diversity in attendance will permeate all MLSE events, and enable a visioning working group at the meeting to develop an inclusive report on the future of machine learning.
“The first MLSE last summer was a great success, providing a new forum for machine learning discussions among scientists and engineers,” said Randall. “It’s very exciting that this grant allows us to expand the event for the next two years by including more students, women, and adding a workshop promoting theoretical foundations, consistent with the goals of TRIAD and IDEaS.”
Investigators:
Award Amount: $200,000
Research: Scaling Up Descriptive Epidemiology and Metabolic Network Models Using Faster Sampling
Santosh Vempala and researchers at the University of Washington are interested in solving a sampling problem that will help researchers spanning many disciplines. Sampling from a given distribution from a space with many attributes is a fundamental problem in computer science. Over the past two decades, practical applications of sampling have proliferated in statistics, networking, biology, differential privacy, and, most notably, machine learning. Sampling is used to evaluate models, as a subroutine for optimization, and more generally for exploring large complex spaces.
The researchers will help develop a toolkit for sampling and evaluate it on real data sets—a large-scale, high-dimensional toolkit for sampling smooth and non-smooth distributions, and a suite of functions that can be computed or estimated using access to samples. It will be developed by working with domain experts in health metrics and systems biology.
Investigators:
Award Amount: $600,000
Transdisciplinary Research Institutes
Georgia Tech’s TRIAD, part of a community of TRIPODS institutes that share expertise and work together, integrates research and education in mathematical, statistical, and algorithmic foundations for data science. TRIAD also hosts focused working groups, national and international workshops, and organized innovation labs, to share data science insights and resources locally and nationally.
“The TRIPODS program, and with it our own TRIAD institute, were established to expand our collective capabilities and accelerate progress,” said Xiaoming Huo, executive director of TRIAD. “Whether it is for education, defining a vision for the future, or pushing the frontiers of research, the new ideas we need come from bridging the boundaries of science, engineering and mathematics.”
]]>Institute for Data Engineering and Science
]]>Although automating these processes improves how quickly decisions are made, it can also exacerbate inequality. Machine learning models find patterns in historical data and use those to predict future patterns. If a certain population had more historical access to some resource (such as mortgages), a machine learning algorithm can identify this trend and try to replicate it in future decisions.
This problem is something School of Computer Science Assistant Professor Jamie Morgenstern thinks about a lot. The entire machine learning pipeline needs to be scrutinized, from data gathering and cleaning to learning the model and deploying it.
Bridging the gap between economics and computer science
This application of game theory to economics has always interested Morgenstern, but it wasn’t until she took a computer science course in her sophomore year at University of Chicago that she found the right outlet for it.
“I liked the implications of the theorems that showed how to solve problems that had some relation to the real world in computationally efficient ways,” she said.
One of these real-world problems is how to think about economics with limited data. Economists tend to believe the world has a perfect data distribution, and people behave rationally in their environments. Computer scientists, however, often operate with a worst-case scenario mindset and recognize people never have a perfect description of their environment — just historical data on it.
“Reasoning about strategic and worst-case behavior are both interesting ways to look at the world and find solutions,” Morgenstern said. “Neither is perfect, but both are useful models in different settings.”
For Morgenstern, machine learning is an ideal bridge between the two schools of thought because it allows generalizations about economic problems despite imperfect information. Morgenstern studied the social impact of machine learning, as well as the impact of social behavior on machine learning, during her Ph.D. at Carnegie Mellon University, where she was advised by renowned theorist Professor Avrim Blum.
Making machine learning fair
Fairness in machine learning is a relatively new area in an already burgeoning field, yet it’s important as the concept gains more popularity.
“The question is how should you make sure you aren’t over-tailoring the information you’re finding from this limited data while still learning?” Morgenstern said.
Morgenstern joined the SCS faculty in Spring 2018, choosing Georgia Tech for its reputation as a theoretical computer science hub ideal for the type of research she does. Her current work focuses on measuring how suboptimal data affects machine learning. A central goal of her work is to ensure machine learning makes the world a more equitable, ethical place rather than further entrenching historical injustices.
“I’ve always cared about social justice and the idea that we should make our systems more equitable and not profitable.”
]]>
SFI is a non-profit independent research center that brings together top academic researchers from the U.S. and abroad. As an external faculty member, Randall will virtually collaborate with other researchers contributing her experience and expertise in theoretical computer science, as well as her research in randomized algorithms and stochastic processes in computer science, discrete mathematics, and statistical physics.
“No aspect of science is going to continue the way it was. New ways of analyzing and integrating the massive datasets being generated in most fields of science and engineering are transforming those disciplines and challenging long-held notions about what can be accomplished,” Randall said in an interview with GT Computing in 2017.
Randall, who is co-director of Georgia Tech's Institute for Data Engineering and Science, is one of 10 external faculty members selected this year by SFI. In all, SFI has more than 100 external faculty members from across a variety of disciplines.
SFI's news release states, “Randall is widely recognized as an expert. Some of her recent work has been dedicated to creating mathematical models for phase transitions and emergent phenomena. Her approaches can be used to understand how organisms flock, how neighborhoods self-segregate, and how swarms of ants or robots can move as one.”
Randall is also the ADVANCE Professor of Computing and an adjunct professor in the School of Mathematics at Georgia Tech.
What started as an attempt to save money by switching the Michigan city’s water supply from the Detroit system to the Flint River in 2014 led to contaminated water and international outcry by 2016. Lead wasn’t coming from the water itself, but the lead service lines that connected Flint homes to the city’s water system.
The problem was no one knew how many homes were affected. Flint hadn’t complied with Environmental Protection Agency standards to keep records of lead service lines. With as many as 50,000 homes potentially needing service line replacements, the city was facing costs up to $250 million. Flint officials were granted only $27 million from the state legislature and an additional $100 million from the U.S. government toward recovery efforts.
Google believed data science could expedite the process of identifying specific homes in need of replacement service lines and donated to the University of Michigan and Flint’s Community Foundation. Abernethy, then an assistant professor at the Ann Arbor campus, and his Data Science Team of students were called in.
Deciding with data
By aggregating a mix of datasets — including pipe material information, parcel records, U.S. census reports, city infrastructure maps, and water samples — Abernethy and his team took advantage of a machine learning technique known as active learning in order to determine which homes were most likely to have lead service lines.
The algorithm performed three processes simultaneously: a statistical model that predicted which portion of the service line was lead, copper, or otherwise; a decision procedure that randomly generated homes for inspection; and a decision-making algorithm that chose which homes should have lines replaced.
Using the predictive model allowed better targeting of resources. Had this method been fully utilized throughout the replacement process, the team estimates that the city could have saved over $10 million — enough money to replace pipes at 2,000 more homes, according Abernethy. The team also created a web-based mobile application to help government offices, residents, and contractors coordinate efforts and update their data in real-time.
Working with a community
Having access to all of this data, as well as numerous algorithmic tools, did not mean that all could be easily implemented on the ground in Flint. Every service line replacement had to be approved by the Flint City Council, and, according to Abernethy, university academics are not always appreciated in towns with struggling economies.
“You write these papers that have these abstract things like, ‘Here’s the way to optimize this whole procedure to save money,’ but what we proposed would take changing large bureaucratic structures and procedures,” Abernethy said. “These interventions are not always welcome.”
Yet the relationship went both ways. For most of the team, this was the first time their research had such a direct impact on a community. Abernethy recalled a city council meeting when a councilwoman asked if her home’s water pipes could be replaced next so that her grandchildren could have safe water to drink.
“It was very touching and moving to have this interaction when we had been thinking of this as an optimization problem,” Abernethy said. “It brought home what we were doing. This is great we’re helping someone like you, but this woman may not have her pipes replaced because of some line in my code. For me, personally, the stakes went up after this conversation.”
Abernethy considers his work in Flint to be one of the most rewarding and challenging of his career.
“If we had picked a project that didn’t have this much public attention, we probably would’ve just written a paper and given up and not done anything useful,” he said. “This was an issue where we had the incentives to go in and do this work very thoughtfully and deeply and really make a difference. I’m really proud of that, and feel very lucky to have had the opportunity to engage in such an important issue.”
Abernethy presents the research in the paper, Active Remediation: The Search for Lead Pipes in Flint, Michigan, coauthored with University of Michigan faculty and students including Assistant Professor of Marketing Eric Schwartz, Alex Chojnacki, Arya Farahi, and Brigham Young University Ph.D. student Jared Webb. The work won the Best Student Paper at the Knowledge Discovery and Data Mining (KDD) conference in London from August 19-23.
]]>
KDD is the world’s premier forum for advancement, education, and adoption of the science of knowledge discovery and data mining and is being held August 19-23. The program includes keynotes, invited talks, workshops, hands-on tutorials, and conventional tutorials.
Faculty and students in the School of Computational Science and Engineering (CSE) and associated research units – Center for Machine Learning, School of Computer Science, and the Institute for Information Security and Privacy, are presenting at this year’s conference.
“It is incredibly exciting to be part of the Georgia Tech team of talented students and faculty, showcasing our diverse research achievements at KDD, ranging from scalable methods for health applications, novel approaches for searching for lead pipes in Flint, to practical defense for artificial intelligence,” said CSE Associate Professor Polo Chau.
Chau will be joined by more than 10 other faculty members and student researchers from Georgia Tech sharing 10 papers in oral, poster, and demo presentations during the five-day event.
In addition to these 10 papers – including one of Chau’s research team’s latest publications, which aims to protect artificial intelligence systems from malicious attacks – Georgia Tech researchers are presenting a conventional tutorial and an all-day workshop. The workshop, Interactive Data Exploration and Analytics, addresses the development of data mining techniques with 13 papers accepted for presentation from various research institutions and groups across the globe.
To help attendees and others interested in the conference, Georgia Tech has developed an interactive data visualization detailing of its research papers, as well as dates, times, and locations for the associated talks.
Below are the titles of Georgia Tech’s research being presented this week along with each paper’s respective authors. Each title is linked to the full paper.
Research Papers:
Workshops and Tutorials:
“The HEAL will provide needed research to inform the national medical imaging policy debate and develop new approaches for improving population health,” said Danny R. Hughes, a Georgia Tech professor of economics and executive director of the Neiman Institute, who will lead the lab. “Drawing on Georgia Tech’s unparalleled strength in interdisciplinary research, the HEAL is uniquely positioned to exploit the vast stores of medical data now available to ensure we move toward a sustainable health care system.”
The center aligns well with Georgia Tech’s core research areas, said President G.P. “Bud” Peterson.
“At Georgia Tech, one of our fundamental missions is help solve society’s most challenging questions,” Peterson said. “Issues of population health, cost, and access to health care are certainly among the most complicated we face. I am certain Georgia Tech’s strong emphasis on data engineering and public policy will provide a firm foundation for Dr. Hughes and the HEAL team.”
The lab will support full-time post-doctorate researchers, graduate research assistants, and affiliated Georgia Tech faculty to produce both methodological and policy-oriented research. A secondary aim of HEAL is to provide training and mentorship to radiologists interested in performing research into health economics and health policy.
Jacqueline Royster, dean of the Ivan Allen College of Liberal Arts, said the center will bring new possibilities for research and collaboration to Georgia Tech and the Neiman Institute.
“We are excited by the opportunities HEAL will bring to both organizations,” she said. “This new lab will benefit from the enormous breadth of top-notch interdisciplinary expertise across Georgia Tech and in the Ivan Allen College of Liberal Arts, from economics, public, policy, and data analysis to artificial intelligence, engineering, and computer science.”
In addition to financial support, the Neiman Institute will provide HEAL researchers access to their extensive data resources, to include large-scale medical claims databases covering millions of U.S. residents.
“This partnership provides a tremendous opportunity to leverage the Neiman Institute’s policy expertise with the analytical capabilities of a world-class engineering institution to address the pressing problems of improving population health, increasing access to medical care, and reducing medical costs.” said Geraldine McGinty, Chair of the American College of Radiology’s Board of Chancellors.
To arrange an interview with a Neiman Institute spokesperson, contact Nichole Gay at (703) 648-1665 or ngay@neimanhpi.org.
]]>Nichole Gay
(703) 648-1665
Lance Wallace
(404) 894-7460
]]>“When you’re doing careful data science, you do care about efficiency, but there are a lot of other values like residents’ voices in the data science process and then in the outcome of what the data gets used for,” said School of Computer Science (SCS) research scientist Amanda Meng.
Meeting a community
For three years, Meng and School of Literature, Media, and Communication Associate Professor Carl DiSalvo have worked with the Westside Atlanta Land Trust, an initiative of the nonprofit, HELP ORG, Inc., that has been trying to establish a community land trust. A community land trust is a nonprofit that owns the land while residents buy the buildings, enabling more accessibility and affordability. Yet before they can become a fully functioning community land trust, they must survey the land to catalog conditions of structures and vacant lots. In March and April 2017, a team of residents collected this data and plotted it on a map, with aid from Meng and SCS master’s student Tom Nguyen.
Having residents collect data is not only empowering for their community, but also leads to more accurate collection. It’s one of the fundamental concepts of Meng’s research with SCS Professor Ellen Zegura and DiSalvo on bringing the sociological concept of care — an attentiveness to the desires and values of the community — to data science.
“The residents know the area,” Meng said. “They are the best at collecting data on what’s changed.”
Working with a community
Collaborating with a community can be challenging for both parties. Researchers must build a relationship so the community trusts them. Meng has been attending community meetings for years and communicates extensively to make sure everyone understands the process.
Yet researchers also must rely on residents to attend data collection meetings, while meeting grant and paper deadlines. Researchers and residents also often have to collaboratively tinker with the data to ensure it’s as clean as possible, from matching street names to what’s in the Fulton County database to checking that each data point corresponds accurately to the map. The effort is worth it, though.
“It’s so rewarding being invited to walk the street with community members and being entrusted with this important goal of capturing changes in property conditions and values,” Meng said.
Meng, along with Zegura and DiSalvo, presented their paper, Care and Practice of Data Science for Social Good, at the new ACM Conference on Computing and Sustainable Societies (COMPASS 2018) held in California from June 20 to 22. The new sustainable society and computing conference was co-organized by Zegura to bring together similar disciplines like computational sustainability and data science for social good.
]]>The 10-week summer program brings college students from across the country to Georgia Tech to use data science research and applications for direct civic and social impact. The National Science Foundation–funded program is now in its fifth year, previously under the name Data Science for Social Good.
This year’s projects addressed gentrification, sustainable transportation, pest control, and environmental monitoring. Each project pairs with local organizations, such as the City of Atlanta and neighborhood planning units, to ensure the work can help the community.
“Community organizations help make sure the kind of problems we’re working on are grounded in reality,” said School of Computer Science (SCS) Professor and program co-director Ellen Zegura. “It’s public problem solving.”
It also lets students see how their data skills can be used outside the classroom.
“The CDS program is unique in that it provides the perfect balance between research and application,” said Michael Koohang, a rising fourth-year student at Middle Georgia State University. “While we conducted formal research during the program, we were also applying our discoveries to tangible pieces of work that had almost immediate impact on local communities.”
For many students, who often come from smaller liberal arts colleges, this was their first opportunity in an environment as large and well-resourced as Tech.
“I knew I was interested in attending graduate school before I came to Tech, but working full-time in a lab, with a mentor and team, has given me invaluable insight about the day-to-day of research,” said Wellesley College rising third-year student Annabel Rothschild.
Tech’s focus on interdisciplinary research also showed students the potential fields they could go into.
“Prior to coming to CDS, I had a lot of difficulty trying to figure out how to combine both my majors, statistical and data sciences and government, into something that excited me,” said Arielle Dror, a rising third-year student at Smith College. “Spending time in a bigger university than my own home institution showed me the exciting world of interdisciplinary research.”
Zegura co-runs the program with Literature, Media, and Communication (LCM) Associate Professor Christopher Le Dantec. They also use the support of faculty mentors: School of Public Policy Assistant Professor Omar Asensio, LCM Associate Professor Carl DiSalvo, LCM Assistant Professor Yanni Loukissas, SCS research scientist Amanda Meng, and School of Civil and Environmental Engineering Associate Professor Kari Watkins.
The four projects included:
Mentors: Carl DiSalvo, Amanda Meng, Ellen Zegura
Problem: Rats are everywhere, but no reliable public data is kept because rat control is outside city jurisdiction and only homeowners can report rats. Without oversight, rats can increase disease, asthma, stress, and cause infrastructure damage.
Solution: Luo and Koohang created an SMS chat bot where residents could report rat sightings via text. They created an interactive map with this data that lets users toggle between layers of code violations to see where rats are. The map can help city officials direct mitigation efforts and provide citizens with a tool to engage the government into action.
Mentors: Yanni Loukissas
Problem: Maps often depict an idealized environment created by the population in power and not everyone’s reality. This team wanted to document and reflect upon the connections and disjunctions between civic data and lived experience in Atlanta.
Solution: They created the Atlanta Map Room in Technology Square Research Building, where everyone can collaborate on large-scale, interpretive maps. Using an app that allows users to select an area of the city to focus on and project it on a piece of paper, users can write their experiences on the map and bring their narrative back to the data. The map allows users to critique data and recognize it may not always tell the full story.
Mentors: Omar Asensio
Problem: Although electric vehicles make up just 2 percent of car sales today, they will be 55 percent in 2050. Despite this boom, charging station experiences are less than accessible.
Solution: This group pulled data from the app Plugshare, where users rate electric vehicle charging stations, to determine how well the current electric vehicle structure serves drivers. They used machine learning to automatically classify all the reviews as having a negative or positive sentiment. Overall roughly 40 percent of drivers have a poor experience at charging stations, a problem that needs to be fixed as the market expands.
Mentors: Christopher Le Dantec and Kari Watkins
Problem: Traffic in Atlanta grows worse every year, but better bike infrastructure can alleviate congestion. Yet the heaviest trafficked routes often have higher pollution, which adversely affects cyclists’ health.
Solution: Students attached low-cost air quality sensors to bikes and ran a series of calibration tests against high-precision sensing equipment. The data will enable a large-scale deployment of bikes to collect air quality data from around the city, determining which routes are too unhealthy for cyclists.
]]>The South Big Data Hub is co-led by Georgia Tech and the University of North Carolina at Chapel Hill. Srinivas Aluru, a College of Computing professor and co-executive director of the Institute for Data Engineering and Science (IDEaS), is a South Hub principal investigator. Renata Rawlings-Goss serves as the South Hub co-executive director for Georgia Tech.
Visit this link to read the full story.
]]>Institute for Data Engineering and Science
jsalazar@gatech.edu
]]>In a new Scientific American article, ISyE Professor Eva Lee describes her efforts as lead analyst on a multidisciplinary team of researchers from Emory, the CDC, and Georgia Tech to predict the immunity of a vaccine without exposing individuals to infection -- including flu, yellow fever, and malaria.
To learn more about the success of the DAMIP model that Lee's team developed, and the role of operations research in developing new vaccines, read more here: http://bit.ly/2lHQ4wA.
]]>Abdallah, who is a Georgia Tech alumnus with master’s and Ph.D. degrees in electrical engineering, conducts research and teaches courses in the general area of systems theory with focus on control and communications systems. He will move into the EVPR role in mid-August, succeeding Stephen E. Cross, who is stepping down after serving as Georgia Tech’s first EVPR for the past eight years.
“Dr. Abdallah has a proven track record as an administrator, scholar and researcher, along with experience collaborating with industry, government and community partners,” said Georgia Tech President G.P. “Bud” Peterson. “As a Tech alumnus who has remained engaged with the Institute, he brings a unique perspective. We’re looking forward to working with him to enhance Georgia Tech’s basic and applied research and maximize economic impact.”
Beyond his service as provost and executive vice president for academic affairs, Abdallah is a professor in the Electrical and Computer Engineering (ECE) Department at the University of New Mexico. He has also served there as acting president, interim president and president, and was the first recipient of the New Mexico ECE department’s Lawton-Ellis Award for combined excellence in teaching, research and student/community involvement. He has received research funding from the National Science Foundation, the Air Force Office of Scientific Research, the U.S. Naval Research Laboratory, other national laboratories and various companies.
“After an academic career of 30 years at the University of New Mexico, I am very excited to be joining one of the premier institutions of higher learning in the world,” said Abdallah. “I look forward to contributing to President Peterson's cabinet, to advocate for and lead the research enterprise at Georgia Tech, and to help ‘create the next’ in research.”
Abdallah is married to Catherine Cooper, a logistics expert who earned her bachelor’s degree in industrial and systems engineering from Georgia Tech. Their twin sons, Carter and Calvin, are incoming Georgia Tech first-year students. Abdallah is fluent in English, French, and Arabic, and serves on Georgia Tech’s ECE Advisory Board. Abdallah is a senior member of IEEE and a recipient of the IEEE Millennium medal.
Christopher Jones, associate vice president for research, has agreed to serve as interim EVPR until Abdallah’s arrival.
Cross plans to step down June 30 to return to his research faculty position in the Georgia Tech Research Institute (GTRI). In welcoming Abdallah, Peterson thanked Cross for his many accomplishments in leading the research program. In addition to dramatic growth in research volume, Cross led the formation of Georgia Tech’s system of 11 interdisciplinary research institutes (IRIs) to coordinate research in areas of specific interest to industrial companies.
“I want to take this opportunity to thank Steve Cross, who has served us well as Georgia Tech’s first EVPR,” said Peterson. “I am personally grateful for his leadership in the development of our 25-year strategic plan, ‘Designing the Future.’ In his EVPR role, he has worked to enhance Georgia Tech’s research impact by bringing together GTRI, the Enterprise Innovation Institute, IRIs and other Institute research, encouraging collaboration, making it easier for government, business, and industry to partner with us, and building innovation neighborhoods around the campus.”
Abdallah’s selection resulted from a national search led by Georgia Tech Provost Rafael L. Bras, who holds the K. Harrison Brown Family Chair.
Research News
Georgia Institute of Technology
177 North Avenue
Atlanta, Georgia 30332-0181 USA
Media Relations Contact: John Toon (404-894-6986) (jtoon@gatech.edu)
]]>
]]>
Research News
(404) 894-6986
]]>jsalazar@gatech.edu
]]>The all-day event featured overview talks and hands-on sessions with this novel hardware, known as field-programmable analog arrays (FPAA). It was led by School of Electrical and Computer Engineering and CRNCH Professor Jennifer Hasler. Several local and industry participants also presented short talks on their research into neuromorphic computing and current challenges in the area.
An FPAA is a configurable chip that uses analog and digital logic to implement algorithms, drastically reducing the power and size of the chip. For certain applications like neuromorphic computing, analog requires fewer transistors to do the same amount of work as digital and produces fewer errors. While analog computing is traditionally a challenging field for computer scientists, hardware like the FPAA and associated high-level and open-source toolsets developed by Hasler’s group allow researchers to create new analog-based algorithms.
“A single programmable and configurable analog chip can compute a large number of things,” Hasler said.
Neuromorphic computers replicate the brain in structure, meaning they are highly parallel, run at lower power, and the fundamental unit computation is very small. The low-power nature of FPAAs makes them ideal for neuromorphic computing. Analog and digital hardware like the FPAA could be pivotal to providing new avenues for computing performance in the post-Moore’s law era, in which the number of transistors on an integrated circuit are no longer expected to double roughly every two years as they have for the past half-century.
“Interesting, novel hardware may be able to solve this post-Moore problem,” said Jeff Young, a research scientist in the School of Computer Science and organizer of the workshop.
The workshop gathered Tech students as well as researchers from prominent national labs such as Sandia, Oak Ridge, Lawrence Livermore, and Pacific Northwest National Laboratory. The speakers’ short research talks expanded on how neuromorphic computing will evolve to tackle post-Moore challenges, including:
As a follow-up to the workshop, the FPAA will be added to the Rogues Gallery, CRNCH’s collection of obscure and unique hardware. CRNCH provides researchers from around the world access to these machines as part of its strategic goal of rethinking high performance computing. When the gallery launched in fall 2017, it had one specialized type of hardware, the Emu Chick, but has since grown to include field-programmable gate arrays (FPGAs), 3D stacked memory devices, and now the FPAA.
]]>
Visit this link to read the full story.
]]>Institute for Data Engineering and Science
jsalazar@gatech.edu
]]>A team of researchers from Georgia Tech and other local universities would study this problem with a grant up to $2 million, dependent on successful completion of milestones, from the Defense Advanced Research Projects Agency’s (DARPA) Lifelong Learning Machines (L2M) program managed by Dr. Hava Siegelmann. DARPA’s goal would be to develop new machine learning approaches that enable systems to learn continually while they operate and apply prior knowledge to new situations.
School of Computer Science Professor Constantine Dovrolis, Georgia Tech Research Institute Senior Research Scientist Zsolt Kira, Georgia State University neuroscience Professor Sarah Pallas, and Emory biology Associate Professor Astrid Prinz are collaborating on the two-year project.
Bringing neural networks to the 21st century
The concept of modeling a computational neural network based on the brain first arose in the 1950s, but it hasn’t evolved much since.
“Obviously, since the ‘50s there’s been a lot of progress in neuroscience, but not a lot of it has translated to machine learning,” Kira said. “Supervised machine learning through neural networks is fundamentally a computer scientist’s translation of a high-level understanding of the brain from the past. But I think there’s a lot we can learn from contemporary neuroscience.”
One of the fundamental problems of machine learning that neuroscience could alleviate is what Dovrolis calls “catastrophic forgetting.” When the artificial neural network learns a new task, it often forgets the previous one.
“Deep learning networks are very different from the brain, both in terms of structure (architecture) and function (dynamics),” Dovrolis said.
Take the brain of a baby. Within the first few years of life, it not only has the ability to learn but also to generalize with very little supervision. Dovrolis believes that machine learning can achieve the same goal but only through a major departure from the currently established machine learning paradigms.
“The brain is really the only example of general intelligence we have,” Dovrolis said. “It makes sense to take that example, identify its fundamental principles, and transfer them to the computational domain.”
Bridging the gap between neuro and computer science
It may make sense, but it’s also controversial. Many computer scientists see the brain as mere hardware, and they prefer to focus instead on more statistical machine learning approaches. This is why this project is so unique: It brings together different ideas from network science, machine learning, evolutionary computing, computational neuroscience, and systems neuroscience — fields that should’ve been working together from the start.
“It’s easier for each field to work by themselves because it’s very comfortable,” Kira said. “But there’s a lot of potential if you actually make the effort to bring people together.”
Yet working with neuroscientists doesn’t just benefit computer scientists. Many neuroscientists believe computing could help with better modeling of biological networks, and ultimately, a deeper understanding of how the brain works.
“Neuroscience can in turn be guided by results from machine learning research that can inform new experiments to deepen our understanding of the brain,” Prinz said.
One of these examples is the flexibility of the brain.
"Neural circuits in the developing brain are highly flexible and adaptable to environmental changes, which endows them with an ability to learn rapidly and to self-repair after damage,” Pallas said.
Adult brains are much less plastic, so one of the neuroscientists’ goals is to uncover the neuronal mechanism that regulates the level of plasticity versus stability in brain circuits. With this, they can harness the mechanism for medical purposes and design machines that can continue learning without forgetting.
Approaching the research
The project should attempt to address five goals of the L2M program:
• Continual learning: The building block of the cortex is a largely invariant structure referred to as a “cortical column.” The function of cortical columns is not known yet, but it seems that they act as associative memories and predictors. The structure of cortical columns suggests that recurrent neural networks could learn incrementally and in an unsupervised manner simply by interacting with the environment.
• Adaptation to new tasks/environments: These cortical columns could interconnect through deep brain-wide hierarchies and nested feedback loops that interact and inform each other, enabling the brain to adapt to different environments with minimal need for re-training.
• Goal-driven perception: At any time, the brain is receiving data from many sensory sources. Hierarchical neural networks could use task-driven inputs to adjust low-level sensory processing and integration dynamically, depending on top-down goal-related signals.
• Selective plasticity: The project should be investigating how the connections and weights between (artificial) neurons could be adjusted when a new task is encountered, without catastrophically forgetting previous tasks. Neuromodulator-driven plasticity and homeostatic plasticity are two biological mechanisms that could be transferred in machine learning to address this problem.
• Monitoring and safety: Researchers would also investigate how to ensure stability and safety, based on the organization of the brain’s autonomic nervous system. Additionally, the safety concern could be further addressed through an “artificial impulse control” system, operating on the same prediction principles as the corresponding cortical system.
This research could allow machine learning to become increasingly adaptive and continually learn, which could have vast applications. A self-driving car could be programmed in the summer, but with these principles, it could learn how to drive in previously untested winter conditions. Siri could be next.
Approved for Public Release, Distribution Unlimited
]]>The technical goal for the new system, dubbed “Gnomon,” is to detect changes in individual computer systems by analyzing suspicious network traffic that appears weeks or months before any evidence of malicious software – or malware – can be identified. As a proof-of-concept, the researchers will work with two major U.S. telecommunication companies and several petabytes of data in basic research aimed at detecting signals of malicious activity on their networks.
Funded by the Defense Advanced Research Projects Agency (DARPA), the four-year award is part of the agency’s Harnessing Autonomy for Countering Cyberadversary Systems (HACCS) program. Beyond rapid detection of infections, the project will also accelerate the cleanup after such infections, creating a clearer pathway in a process known as remediation.
“A compromise becomes a breach only if the original infection remains undetected long enough for the adversaries to do damage,” said Manos Antonakakis, an assistant professor in Georgia Tech’s School of Electrical and Computer Engineering and the project’s co-principal investigator. “If you look at the major breaches that have occurred, you see that the adversaries were in the systems for months. We want to identify them in a matter of hours to contain the infection before any real damage can be done.”
The new techniques to be developed will address the realization that network attacks cannot be completely blocked by existing defenses and malware-based detection systems. Dynamic intelligence will be a key feature of the system, with the intent of creating a continuously-updated dossier of every address in IPv4 space.
“Gnomon will search for illicit behavior in computer systems and network signals that indicate the start of an infection,” said Michael Farrell, chief strategist at the Georgia Tech Research Institute (GTRI), and the principal investigator on the program. “We’ll use our experience with taking down botnets – networks of infected computers – to accelerate the detection and remediation process. It’s imperative to evolve our view of the internetwork infrastructure at the same pace that the threat evolves.”
To protect millions of computers on the networks of the two companies, the researchers must find ways to identify troubling behavior on individual IP addresses without endangering the privacy of individuals. Among the signs of trouble are communications with network locations known to house malicious activity. Such communication is necessary for malicious groups to control computers that have been compromised, and to move data stolen from them.
“If you know where the infecting groups are located, you can very easily exclude most of the benign activities occurring on the network,” Antonakakis said. “We need to be able to identify what has changed in computers throughout the network, understand why the change has happened, and determine whether that change can be attributed to benign or malicious activity. This is a groundbreaking new approach to network security that will require tremendous computing power and infrastructure.”
Ever since the first viruses hit computers in the 1980s, cybersecurity has seen rapid evolution of detection and attack tactics. The success of Gnomon will likely drive adversaries to new attack techniques that may be more complex – and expensive – than existing activities. Making cyberattacks more costly to launch may reduce the profit from such activities, making them less attractive.
“If we can clean up our networks faster and more efficiently, that will increase the cost of the attack, making the adversaries work harder,” Antonakakis said. “If you raise the cost of an attack, the return on investment becomes smaller, while the risk of getting identified becomes higher. We would like to make the business of an attack so unprofitable and so risky for the adversaries that it will not make sense for them to conduct major operations in our networks.”
Success in developing new techniques with the first two telecommunication companies could open the door for scaling up Gnomon to other large networks in industry – and to U.S. government systems.
“Not only will deployment have an obvious benefit of improved hygiene for a significant portion of the U.S. internet infrastructure, but the public-private partnership will allow us to provide valuable feedback throughout the HACCS program on the sort of prototypes that will be necessary to have true business and mission impact in the real world,” Farrell said. “The goals are very ambitious, but if we’re successful, we’ll be able to close the gap between an infection and remediation.”
This program is the latest interdisciplinary research collaboration in cybersecurity at Georgia Tech, orchestrated by the Institute for Information Security & Privacy (IISP). In addition to the School of Electrical and Computer Engineering and GTRI, the project will include Professor Brian Kennedy from Georgia Tech’s School of Physics.
Attribution of malicious cyber activity is an established research thrust at Georgia Tech, and this new contract builds on the early success of another Department of Defense (DoD) sponsored program to enhance attribution. The “Rhamnousia” program is now a $25.3 million contract being led by the same research team of Farrell and Antonakakis.
This material is based upon work supported by the Defense Advanced Research Projects Agency (DARPA) under contract number HR001118C0057. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the Defense Advanced Research Projects Agency (DARPA).
Research News
Georgia Institute of Technology
177 North Avenue
Atlanta, Georgia 30332-0181 USA
Media Relations Contact: John Toon (404-894-6986) (jtoon@gatech.edu).
Writer: John Toon
]]>Research News
(404) 894-6986
]]>The White House Office for Science and Technology Policy (OSTP) is hosting the daylong meeting, which executives from 35 major U.S. companies including Facebook, Amazon, Google, Intel, NVIDIA, and Oracle are expected to attend.
The focus of the meeting is AI research and development, workforce, and regulations.
Georgia Tech College of Computing Executive Associate Dean Charles Isbell, along with a small group of his peers from academia, has been invited to participate in the discussions.
“There’s not a sector in the U.S. economy that isn’t being impacted by AI,” said Isbell. “But that doesn’t mean that what makes regulatory sense for, say the auto industry, will necessarily make sense for pharmaceuticals.”
He added, “The U.S. has a real opportunity to take the lead and set the standard for the ethical, human-centered development of AI – regardless of the sector – to deliver on the promise of this transformative technology in ways that we have yet to imagine.”
Representatives from the Pentagon and the U.S. Departments of Agriculture, Commerce, Energy, Health and Human Services, Labor, and Transportation are also slated to participate in the White House AI meeting.
Isbell was in Washington earlier this year testifying before the House Oversight Committee’s Subcommittee on Information Technology about the national and global impact of AI and machine learning.
]]>Woodruff School Professor Surya Kalidindi has been named a recipient of the 2018 Department of Defense Vannevar Bush Faculty Fellowship. Kalidindi, the first Georgia Tech faculty member to receive the fellowship, joins a group of 45 current Vannevar Bush Faculty Fellows who are sponsored by the DoD to conduct foundational research in core science and engineering disciplines underpinning future DoD capabilities.
Read the entire article here.
]]>
Several objectives were accomplished, including beginning a vibrant discussion that identified specific areas of mutual interest between corporate attendees and IDEaS-affiliated researchers. Follow-up activities are being planned to increase industry engagement with IDEaS.
Co-Executive Director Srinivas Aluru kicked off the event, and Georgia Tech’s Executive Vice President of Research Stephen E. Cross welcomed the day’s participants with a view of interdisciplinary research across Georgia Tech. Several presentations by industry and researchers served to share expertise, challenges, and perspectives, and to begin a conversation between participants.
The day opened with introductory presentations. Dr. Aluru addressed the structure and goals of the data science institute at Georgia Tech. “Georgia Tech historically prides itself in applied innovation and being an asset to industry,” he said. “In the area of data science in particular, most of the big data problems originate in industry, making academia-industry collaboration not only preferable, but essential.”
Director of Industry Partnerships Renata Rawlings-Goss provided an overview of opportunities for industry to become involved via the IDEaS Industry Alliance Program. Of the event she said, “Industry participation in the data science ecosystem here at Georgia Tech is truly a win-win for both sides. We have top ranked programs, faculty, and research here at Tech, but we also strive to be good partners in the dynamic industry analytics community here in Atlanta and nationally. Industry Day is one way we listen and connect with Industry partners and local companies.” Dr. Rawlings-Goss asked interested industry and faculty researchers to get involved by contacting her.
IDEaS is also closely allied with the South Big Data Hub, with Aluru (as the hub PI) and Rawlings-Goss (as the hub Co-Executive Director) serving in leadership roles of both entities. Although this is the first Data Science Industry Day, IDEaS has run workshops dedicated to bridging industry and academia in the past, in disciplines like materials and manufacturing, analytics and machine learning, precision medicine, and IoT for smart and connected cities and campuses.
Several themed panels took place in the late morning and afternoon:
Dana Randall, co-executive director of IDEaS, moderated the session on Centers and Educational Programs. Faculty and staff from Georgia Tech highlighted several opportunities:
David Sherrill, an IDEaS associate director, moderated the session on AI and Deep Learning. Panel participants included
Deirdre Shoemaker, an IDEaS associate director, moderated the session on Graph Analytics and Anomaly Detection
Marilyn Wolf, an IDEaS associate director, moderated the session on Unstructured Data Analytics (Video, Text, Image)
Between panel presentations, multiple “rapid fire” industry sessions encouraged industry guests to introduce themselves, describe their company and work, and how they envision interacting with Georgia Tech.
The evening concluded with poster presentations from nearly thirty students in the event’s three theme areas of AI and machine learning, graph analytics and anomaly detection, and unstructured data composed of text, image, and video.
The event was planned by Shkina Halbert of IDEaS.
IDEaS extends its sincere thanks to Accenture, Microsoft, Microsoft Research, United HealthCare, and Equifax for sponsoring our event.
IDEaS at Georgia Tech
The Institute for Data Engineering and Science unifies data science researchers and resources spanning all disciplines throughout Georgia Tech to take on grand challenges in data science. It strategically builds collaborations and supporting resources to stimulate foundational research in areas such as machine learning, high-performance computing, and algorithms and optimization. It identifies and unites researchers to pursue collaborative and ambitious funding opportunities, to drive research, and to evolve and promote data science education. IDEaS provides an accessible and stable means of navigating the vast landscape of data science research and opportunities internally, and externally as it connects to industry and other partners. Join the conversation by contacting a member of the IDEaS team and learn more at ideas.gatech.edu.
Interdisciplinary Research Institutes at Georgia Tech
The twelve Interdisciplinary Research Institutes (IRIs) of Georgia Tech are responsible for bringing together a mix of researchers – spanning colleges, departments, and individual labs – around a single core research area. IRIs also connect a large portfolio of basic and applied research programs, support world-class research facilities and laboratories, engage Georgia Tech students, and collaborate with government and industry research partners.
]]>[WATCH: CRNCH Time at Georgia Tech for Rebooting Computing Architecture]
The National Academies of Science, Engineering, and Medicine’s National Materials and Manufacturing Board (NMMB) Spring 2018 Meeting will address this question with its session Moore’s Law: Where We’ve Been and Where We Are Going in an Age at the End of Scaling on March 20 at the National Academies in Washington, D.C.
Tom Conte, a School of Computer Science professor and co-director of the Post-Moore–focused Center for Research into Novel Computing Hierarchies (CRNCH), joins the event to offer his perspective on technology and alternative paths.
The session is part of a larger daylong event during which the NMMB hopes to identify areas that could impact government policymakers and program managers. Specialized electronics are integral to the defense industry, which needs reliable products that can be produced at low cost.
The post-Moore era is ripe for opportunities in technological innovation but could also present challenges in policy and economics. During this session, industry, academia, and government experts will discuss what the future of computing faces—from how to prolong the life of transistors to exploring alternative materials and architectures to push it forward.
The session will be divided into three panels on:
The Institute for Data Engineering and Science (IDEaS) teamed with Correlation One and sponsors Citadel LLC and Citadel Securities to host a Datathon, or Data Open, in the Klaus Advanced Computing Building on Feb. 3. Faculty members Polo Chau (School of Computational Science and Engineering), Guillermo Goldsztein (School of Mathematics), and Dana Randall (School of Computer Science and co-executive director of IDEaS) served as judges. Nearly 500 students applied to participate, and the top 100 were invited to compete. At the competition, aspiring data scientists analyzed a problem statement and data sets related to the future of urban environments and "smart" cities. Students competed in teams of four for a $25,000 cash prize.
Read the full story here.
The Georgia Tech College of Computing continued its climb up the U.S. News and World Report rankings of graduate computer science programs, rising one spot to No. 8 in the 2018 rankings that were released March 20.
The new position represents Georgia Tech’s second jump in the last three CS rankings, all released since 2012, and is the highest U.S. News has ever ranked the College of Computing.
“We are thrilled but not surprised at this latest recognition of the work we’re doing in GT Computing,” said Zvi Galil, John P. Imlay Jr. Dean of Computing.
“I attribute this to our visible leadership in computing education and research, to the fact that we are now the largest computing program in the United States counting both faculty and students–and likely number 2 in terms of faculty size–and to the global impact we are having both through our research and the work of our thousands of alumni.”
U.S. News ranks computer science programs through a reputational survey. With our average score of 4.4, Georgia Tech has now tied with Princeton and one spot ahead of No. 10 University of Texas-Austin. In the 2018 rankings, Georgia Tech rose in both points and ranking—and was the only Top 10 program to rise in either.
The College also achieved rankings in the following specialties:
Coincidentally, the No. 8 overall ranking matches the spot Georgia Tech earned in last year’s Times Higher Education/Wall Street Journal rankings, when the College was named the No. 8 program in the world.
“Over the past several years,” Galil continued, “we have made deliberate, strategic investments of time and treasure, both in research and education, and this recognition is one of the fruits of those efforts.”
The College of Computing was just one of many Georgia Tech programs to place highly in the 2018 rankings.
The College of Engineering also ranked No. 8 (No. 4 among public universities), and all 11 of its programs ranked in the top 10. In the College of Sciences, Chemistry jumped four to No. 20, Math moved up two to No. 26, Physics moved up one to No. 28, Earth Sciences moved up four to No. 38, and Biology moved up one to No. 54. Within mathematics, the discrete math/combinatorics specialty had Georgia Tech at No. 2, up two positions.
[READ: U.S. News and World Report 2019 Graduate School Rankings]
The Scheller College of Business full-time MBA program moved up one to No. 28, and its part-time MBA moved up five to No. 25. Scheller was also ranked in the following specialties:
In the Ivan Allen College of Liberal Arts, the Sam Nunn School of Public Policy moved up two to No. 43 overall with the Information and Technology Management specialty remaining at No. 2, Public Policy Analysis debuting at No. 20 and the Environmental Policy and Management specialty debuting at No. 12.
]]>The morning involved hearing about the opportunities and challenges related to precision medicine and data science from Georgia Tech’s and Emory’s points of view, with Srinivas Aluru, Krish Roy, Jon Duke, Greg Gibson, and May Wang delivering for Georgia Tech and Lance Waller, Mike Zwick (presented by Lance Waller), Vicki Hertzberg and Arshed Quyyumi delivering for Emory. The afternoon was devoted to break-out sessions to identify critical needs and paths forward. Several action items were identified.
Event participants and affiliations:
“I am excited by these hearings,” said Isbell in his testimony. “Advances in AI are central to our economic and social future.”
Isbell – an affiliated faculty member with The Center for Machine Learning at Georgia Tech (ML@GT) – was invited to present committee members with testimony on the “game-changing” nature of AI and to explore the potential impact of additional industry regulation.
The committee requested that Isbell – along with expert witnesses from the Allen Institute for Artificial Intelligence, Intel, and Nvidia – unpack the “potential for artificial intelligence and machine learning to transform the world around us and how we might collectively best respond to its potential.”
“By leveraging computing and data, we are moving from robots that assemble our cars to cars that almost drive themselves,” said Isbell. “One can be skeptical – as I am – that we will in the near future create AI that is as capable as humans are in performing a wide variety of the sort of general tasks that humans grapple with every day, simultaneously.
"But it does seem that we are making strong progress toward being able to solve a lot of very hard individual tasks as well as humans. We may not replace all three million truck drivers and taxi cab drivers, nor all three million cashiers in the United States, but we will increasingly replace many of them. We may soon trust the x-ray machine itself to tell us if we have a tumor, as much as we trust a doctor. We may not automate away intelligence analysts, but AI will shape and change their analyses.”
The ML@GT research center is a key player in the type of interdisciplinary exploration required to apply machine learning to the kinds of autonomous systems (like self-driving cars and manufacturing robots) explored in the Oversight Committee hearing. It is also pushing forward on less obvious applications such as health care, data analytics and data mining, and applications in biology, economics and finance, and social networks.
Isbell noted in his testimony the importance of transparency in algorithm development and the creation of machines that “make decisions with humans” to help us accomplish tasks more effectively. He also stressed that, in order to both help people collaborate more efficiently with AI assistance and enable better human oversight, education will play a larger role in a technologically assisted future.
“To understand such machines, much less create them, we should strive for citizens to not only be literate but to be compurate; that is, they must understand computing and computational thinking and how it fits into problem-solving.”
Isbell is also one of the minds behind Georgia Tech’s new Constellations Center for Equity in Computing, a center based in the College of Computing and serving as a hub for research and advocacy at the intersection of educational equity and computing. Accordingly, Isbell’s testimony stressed the equally significant roles of education and research in the future of the burgeoning fields of machine learning and AI.
“The issues that are being raised here can be addressed with thoughtful support for robust funding in basic research in artificial intelligence—including research in how to engage in education; support for that education throughout the pipeline from K-12 and beyond, including post-graduate retraining; and in developing standards for the proper use of intelligent systems,” said Isbell.
Additional invited expert witnesses for the House Oversight Committee included Oren Etzioni from the Allen Institute for Artificial Intelligence, Amir Khosrowshahi from Intel, and Ian Buck from NVIDIA.
To watch the full hearing or read more about the topics discussed, visit the Oversight & Government Reform website.
]]>“NGS refers to a suite of technologies that enable high-throughput and inexpensive DNA sequencing. Their origin dates back to just a decade ago, but these technologies now underpin all of modern genomics research,” said Aluru.
Aluru’s group is already a pioneer in developing bioinformatics methods for NGS, including a large NSF and National Institutes of Health big data project that came to fruition during the initial round of federal investments in big data.
The two new projects are collectively dedicated to improving the quality and speed of NGS analyses. The first project explores the development of new algorithms to directly tackle errors made by NGS machines through the development of approximate matching algorithms. Previous methods are either time consuming and prohibitive to use on large data, or depend on heuristic techniques with no quality or performance guarantees.
“The new research takes advantage of limited errors made by NGS machines, and develops algorithms that can tackle a bounded number of errors efficiently. It continues a line of investigation initiated through collaboration with late [College of Computing] professor Apostolico,” he said.
The second project will conduct comprehensive experiments on reproducibility, and will evaluate key methods and software products scientists currently use to analyze NGS data. It will benefit the research community by:
As precursor, a paper authored by Aluru and three CSE graduate students was recognized at Supercomputing 2016 as the first selected paper by ACM SIGHPC under the scientific repeatability, replicability, and reproducibility initiative. The new NSF project will also involve nearly 20 undergraduate students, who will learn about research integrity and reproducibility issues.
Both projects impact a wide range of applications: from medical research to evolution, as well as influence work outside of biology in text matching and information retrieval.
]]>Over the years, advances in machine learning have resulted in more complex, and more powerful, applications in information visualization. As a consequence, machine learning techniques to achieve specific insights from data have also gotten more complicated. Most require data science degrees or some formal data science training in order to use the tools that are being built.
Thus, the gap between subject matter experts – international politics majors, historians, biology experts, or climatologists, for example – and the complexity of the machine learning tools used to contextualize data will continue to grow.
“Often, these experts have a wealth of knowledge about things like international affairs or cybersecurity, but they don’t have a wealth of knowledge of what it means to use machine learning model X, Y, or Z,” said Alex Endert, an assistant professor in the School of Interactive Computing at Georgia Tech, one of the four collaborators on the project.
Currently, tools to adjust parameters on the data consist of buttons, control panels, dropdown menus and sliders, knobs and fields to adjust values, direct manipulations to define a machine learning model and letting it achieve the desired data.
This is less intuitive for non-data scientists, so the aim for the researchers is to move the user interaction into the visual space. Users could adjust the data within a scatter plot, for example, by zooming or panning, coloring items or generally demonstrating areas of interest inside the data. Then, they could infer how those parameters should change as a result of the exploration of the data.
“If we are successful, we have the chance to bring data analysis to the public,” said primary investigator Remco Chang, an associate professor in the Tufts University Department of Computer Science. “But to get there, we will need to allow the end users to be able to intuitively ask questions about their data that can be formalized and executed in machine learning. We need to allow the user to make sense of the complex results from machine learning and help contextualize the results in the user’s domain.”
The grant, which took effect earlier this year, will fund four years of research. Other participants are Georgia Tech School of Interactive Computing Professor John Stasko, and University of Wisconsin Department of Computer Science Professor Michael Gleicher.
DARPA’s D3M program aims to develop automated model discovery systems that enable users with subject matter expertise but no data science background to create empirical models of real, complex processes. Automated model discovery systems developed by the D3M program will be tested on real-world problems that will progressively get harder during the course of the program. Toward the end of the program, D3M will target problems that are both unsolved and underspecified in terms of data and instances of outcomes available for modeling.
]]>Communications Officer
]]>“It’s not that simple in climate,” said Annalisa Bracco, a professor in Georgia Tech’s School of Earth and Atmospheric Sciences. “Even weak connections between very different regions on the globe may result from an underlying physical phenomenon. Imposing thresholds and throwing out weak connections would halt everything. Instead, a climate scientist’s expertise is the key step to finding commonalities across very different data sets or fields to explore how robust they are.”
And with millions of data points spread out around the globe, Bracco said current models rely too much on human expertise to make sense of the output. She and her colleagues wanted to develop a methodology that depends more on actual data rather than a researcher’s interpretation.
That’s why the Georgia Tech team has developed a new way of mining data from climate data sets that is more self-contained than traditional tools. The methodology brings out commonalities of data sets without as much expertise from the user, allowing scientists to trust the data and get more robust — and transparent — results.
The methodology is open source and currently available to scientists around the world. The Georgia Tech researchers are already using it to explore sea surface temperature and cloud field data, two aspects that profoundly affect the planet’s climate.
“There are so many factors — cloud data, aerosols and wind fields, for example — that interact to generate climate and drive climate change,” said Athanasios Nenes, another College of Sciences climate professor on the project. “Depending on the model aspect you focus on, they can reproduce climate features effectively — or not at all. Sometimes it is very hard to tell if one model is really better than another or if it predicts climate for the right reasons.”
Nenes says the Georgia Tech methodology looks at everything in a more robust way, breaking the bottleneck that is typical of other model evaluation and analysis algorithms. The methodology, he says, can be used for observations, and scientists don’t need to know anything about computer code and models.
“The methodology reduces the complexity of millions of data points to the bare essentials —sometimes as few as 10 regions that interact with each other,” said Nenes. “We need to have tools that reduce the complexity of model output to understand them better and evaluate if they are providing the correct results for the right reasons.”
To develop the methodology, the climate scientists partnered with Constantine Dovrolis and other data scientists in Georgia Tech’s College of Computing. Dovrolis said it’s exciting to apply algorithmic and computational thinking in problems that affect everyone in major ways, such as global warming.”
“Climate science is a ‘data-heavy’ discipline with many intellectually interesting questions that can benefit from computational modeling and prediction,” said Dovrolis, a professor in the School of Computer Science, “Cross-disciplinary collaborations are challenging at first — every discipline has its own language, preferred approach and research culture — but they can be quite rewarding at the end.”
The paper, “Advancing climate science with knowledge-discovery through data mining,” is published in Climate and Atmospheric Science, a Nature journal.
The development of the methodology was supported by the U.S. Department of Energy (grant DE-SC0007143) and the National Science Foundation (grant DMS-1049095). Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the sponsors.
]]>Constellations is dedicated to challenging and improving the national computer science (CS) educational ecosystem through the provision of curricular content, educational policy assessment, and development of strategic institutional partnerships. According to Senior Director Kamau Bobb, democratizing computing requires a “real reckoning with the race and class divisions of contemporary American life.”
The central challenge facing the United States in computing education is that there simply are not enough K-12 teachers or university faculty at any level to meet the demand of students or the needs of industry. In Georgia, there are approximately 519,000 high school students. In the 2016-17 school year less than 10 percent, only 43,000 students, enrolled in CS courses of any kind. Of the approximately 600 teachers instructing those students, only 33 of them were certified by the state professional standards commission.
However, like any limited and valuable commodity in American life, access is strongly correlated with race, class, and privilege. Black students, for example, comprise 13 percent of all students taking CS courses, even though they represent 37 percent of all public high school students in the state. This pattern is replicated in many states across the country.
“We’re trying to build an equitable infrastructure for CS education within a national educational system that is tragically inequitable. We see this as an issue of educational justice,” said Bobb, who comes to Georgia Tech from the National Science Foundation (NSF), where he served as a program officer in NSF’s Directorate for Computer & Information Science & Engineering.
“As computing emerges from under the STEM umbrella as a national priority and a pathway to power and social mobility in America, there is a tremendous amount of collaboration, development, and advocacy to do. Georgia Tech is putting the full weight of its research capacity and programmatic innovation into this effort.”
Bobb is joined at Constellations by Lien Diaz, who serves as director of educational innovation and leadership. Diaz previously was with The College Board as director of curriculum, instruction, and assessment. In a partnership with the NSF, she led the development of the new Advanced Placement (AP) Computer Science Principles course that launched in the fall of 2016.
“For decades now, computing fields have lacked female representation and racial diversity in the workforce,” Diaz said. “When I was at AP, the goals behind developing CS Principles are quite similar to what we’re trying to do at Constellations: We want to show how relevant computing is, no matter where you live, or what your gender or the color of your skin is. We want to dispel the myth that computing is only for a certain population. It’s for everyone, and computing skills will be too critical to everyone in our future society for us to simply accept the current disparities in access to computing education.”
Constellations comes at a critical time, as the technology industry is being forced to reckon with calls for equity as never before. Like Diaz, Bobb brings critical experience in addressing these very issues. In his role at NSF, he helped construct a national research agenda intended to accomplish many of Constellations major goals, which include:
“The lack of diversity in computing is well-documented and represents a real threat to the nation’s political, economic, and cultural standing in the world,” said Charles Isbell, executive associate dean in the College of Computing.
“Many university faculty members have devoted their research energy to understanding the most effective methods for computing education and increasing access for all students. The College’s ongoing work to offer Georgia Tech CS courses and support to Atlanta Public Schools (APS) and other districts across the state has deepened our grass-roots understanding of the barriers to equity.”
That ongoing work with APS is just one of the reasons why Georgia Tech provides the perfect home for Constellations. College of Computing faculty members Betsy DiSalvo, Barb Ericson, and Mark Guzdial are nationally recognized leaders in the study and design of effective CS educational programs and the systemic barriers to equity in CS education. The College itself, through its Office of Outreach, Enrollment, and Community and the work of Assistant Dean Cedric Stallworth, has a long history of computing outreach and teacher-training programs.
“The Constellations Center aims to highlight Atlanta’s diverse communities, promote the concealed talent that exists to contribute to a growing technology-driven workforce, and diminish racial and gender divides in American social and tech spaces,” Bobb said. “Computing is not just a matter of education, it is an element of a fair and just democracy.”
]]>404-894-7253
]]>The analysis – released to coincide with National Computer Science Education Week – examines College Board data on 2017 national results of the Advanced Placement Computer Science (AP CS) A exam and the recently introduced AP CS Principles (P) exam.
According to Barbara Ericson, Georgia Tech research scientist and author of the analysis, the introduction this year of a new AP CS P course and exam contributed to the increases.
“This is exactly what we hoped for. The CS principles course is on par with a college-level intro course for non-CS majors, so it is more accessible to more people,” said Ericson.
Officials had estimated nearly 20,000 AP CS P exams would be taken this year. However, Ericson said the actual number topped 40,000.
“Although overall growth in female and minority participation in the AP CS A exam was relatively flat this year, we’re hopeful that the introduction of the P exam will help swell A exam participation rates in the next few years.”
Despite marginal growth among underrepresented students, overall participation in the AP CS A exam grew by 11.2 percent year-over-year in 2017. A record 60,519 U.S. high school students took the exam with an overall pass rate of 61.8 percent, up more than a percentage point from the previous year.
“It’s great to see growth across the board, but there’s still a long way to go before AP computer science is as available in U.S. classrooms as, say, AP Physics or Calculus,” said Ericson.
More than 170,000 students took the AP Physics 1 exam this year, while more than 316,000 took the AP Calculus AB exam.
“Increasing the penetration rate for CS is critically important to the U.S. remaining competitive in an international marketplace increasingly reliant on computing and hungry for highly skilled and well-trained computer science talent,” said Ericson.
Despite some positive movement this year, the analysis shows that more needs to be done to improve outcomes for female and underrepresented students. The CS A exam had the lowest participation rate among female students than any other AP exam this year.
The pass rate nationally for women taking the AP CS A exam was 64 percent, which was lower than males by 3.8 percent. The pass rate for African American females was 26.1 percent. In 12 states, no African American females took the exam.
Comprising just 16 percent of all exam takers, the pass rate for underrepresented students was 44 percent.
“Many of these students have less exposure and less confidence in their ability to do computer science,” Ericson said. “They also encounter negative stereotypes about computer science, either people telling them CS is not for them or that CS is boring, antisocial, and doesn’t help anybody.”
In its initial year, 43,780 AP CS Principles exams were given following the 2016 introduction of the new course offering.
Female students represented 30 percent of AP CS P exam takers, while 26 percent were from underrepresented groups, which include American Indian/Alaska Native, Black, Hispanic/Latino, and Native Hawaiian/Other Pacific Islanders. This was the second highest rate of underrepresented minority participation for all STEM (science, technology, engineering, and math) AP exams.
Overall, the pass rate for the AP CS P exam was 75.1 percent. However, the same disparity that exists between majority and minority outcomes on the AP CS A exam is also evident in the P exam.
The pass rate for women was 70 percent, lower than their male counterparts by 6 percent. At 53 percent, the pass rate for underrepresented students was nearly 30 percent lower than the majority on the CS principles exam.
“The goal of the CS P course is to provide an engaging and rigorous introduction to computer science to a broad diversity of students. While the CS P exam takers are more diverse than those who took the CS A exam, the percentage of females is still low and under-represented students have a lower pass rate on both exams," said Ericson.
“More work needs to be done to recruit and support females and minorities in computer science. Hopefully, as a nation, we can build on the initial success of CS P to truly offer computer science for all students.”
]]>MLH is an international student hackathon league. Participating in sanctioned events throughout North America, student teams from across campus collectively competed in most of the 190 events held between August 2016 and July 2017. In all, more than 3,000 university and high school students participated in the North American MLH events.
Georgia Tech rose to the top of the MLH standings this season by earning the highest number of merit points and the second highest number of participation points. Combined, these stats put them ahead of Rutgers University in second place and the University of Waterloo in third place.
Merit points are earned by submitting winning entries at sanctioned MLH hackathons. Teams are awarded participation points for sending teams to compete in events that require travel.
For the first place finish, Georgia Tech received the MLH Season Trophy and $50,000 in equipment from competition sponsors Dell and Windows. HackGT – Georgia Tech’s student hackathon and computer science outreach organization – accepted the prizes on behalf of all Georgia Tech’s student hackathon teams.
“Winning the 2017 MLH Hackathon Season award is not solely an ode to the computer science outreach of HackGT, but to all of the Georgia Tech students who excelled at MLH sponsored Hackathons,” said Jordan Madison, HackGT director of communications and third-year computer science major.
“With the $50,000 in equipment, we will grow the pool of resources that we provide to our attendees; therefore, giving them more opportunities to bring even more ideas to fruition.”
Georgia Tech’s star has been on the rise in MLH competition. In 2013, Tech ranked 16th. In Spring 2016, teams rose to 6th place before earning the top spot in the 2017 season.
This upward trajectory is in no small part attributable to HackGT. Led by Shehmeer Jiwani (CS ’16) and Pavleen Thukral (CS ’16), Georgia Tech students started attending MLH hackathons around 2013. This resulted in the founding of the HackGT student club in 2014.
“HackGT’s mission is to empower the Georgia Tech community and others to utilize computer science as a tool to bring their ideas to life. Through our events, we give attendees the space and resources needed to achieve this goal and hopefully inspire them to continue to utilize computer science in their perspective fields of study,” said Madison.
The growing interest in hackathons on campus led Jiwani and Thukral to organize Georgia Tech’s first hackathon – known as Hackatech – which has since evolved into HackGT. The organization hosted its fourth hackathon, HackGT, in October. HackGT4: New Heights attracted more than 1,000 hackers from across North America to the three-day event.
]]>
“An entirely new type of workforce is needed for the 21-century. One that will require data-enabled talent for jobs across industries and government, as well as for future scientific discovery,” said Rawlings-Goss “That is why we are partnering with sponsors and people spanning many disciplines and roles to make sure that the discussion of data science remains as broad as it needs to be. To achieve the talent pool needed for continued U.S. growth and competitiveness in this new data economy, we must break open and structure education with access for all. Programs must include minorities, reach low-income future students, and pair different institution types like two-year teaching institutions with universities.”
Rawlings-Goss also heads the education working group within the Big Data Hub consisting of a nationwide network that tackles tough problems like these. And she wants everyone to join the conversation. “Data science is something that can be brought into nearly every discipline; the idea that data science resides in only a few traditional math and computer science disciplines is part of the challenge we are trying to overcome in education,” she said.
National Science Foundation Program Manager Stephanie August, who attended the event said, “NSF is involved because we want to get at the missing voices. We need these participants who are essential but mostly absent and unheard in the movement to develop data science programs.”
Many research universities are developing comprehensive data science programs at the undergraduate and graduate levels. As the movement to develop data science programs grows, a gap is forming that separates research institutions from primarily undergraduate-focused institutions, including community colleges and several minority-serving institutions.
Specific issues discussed included access to data, critical thinking, designing curriculum and assessment, data literacy, diversity, ethics, resources and staffing, building collaborations, and the pipeline to higher education from K-12. Participants also addressed training data science practitioners, and translational data science, or the application of data science principles, techniques, and technologies to scientific problems impacting human or societal welfare.
Recent education-enabling projects were showcased at the event. Aleksandr Blekh of Georgia Tech’s College of Engineering introduced participants to his work with JupyterHub, building executable textbooks in support of data science education. The South Hub recently partnered with U.C. Berkeley and completed an installation and demo that included 60 faculty members and deans from across the country interested in using the tool to expand their data science capacity.
The workshop agenda moved from awareness of existing programs, practice and challenges, specific topics and stakeholders, to actions needed to create the vision, curricula, programs, and opportunities of the future. It included discussion of what is important for the future of technology and society in the United States, and improving access to essential and economically viable jobs, especially for minorities and low-income students and workers.
Tasha Inniss, director of education and industry outreach at the nonprofit professional society INFORMS said, “We have all gone to conferences that attempt to discuss these issues and policy without having the full representation of the necessary stakeholders. This workshop, fortunately, has been different, because it includes participation of faculty from minority-serving institutions and community colleges as well as representatives from industry and non-profit organizations. Since INFORMS has members who are analytics and operations research professionals and students, I am particularly interested because it fits nicely with our analytics education goals.
Mary Rudis, a mathematics instructor from Bates College agreed. “This is the first event I have been to that is constructed from the ground up with the right mix, capable of producing a real outcome. The blend of participants in terms of diversity, institutional types, roles, and expertise reflects the very nature of data science. It is multidisciplinary, transdisciplinary, with complex issues that require input from many points of view.”
Rawlings-Goss, Iniss, Rudis, and more than a dozen other authors will continue to work together after the event to develop a report including plans and schedules to convert their vision to reality. Findings will be presented at the National Academies of Science in Washington DC in early December.
This workshop, “Negotiating the Digital and Data Divide,” is part of the “Keeping Data Science Broad” series created by Rawlings-Goss. Other activities include webinars and presentations to garner community input into pathways for keeping data science as a discipline broadly inclusive. The growing community seeks input from data science programs in any region across the nation, either traditional or alternative, and from a range of institution types. Two webinars leading up to the workshop explored the future of data science education and workforce at institutions of higher learning that are primarily teaching-focused. A webinar after this workshop will be announced to report on its outcomes and next steps.
The Keeping Data Science Broad series is co-sponsored by the National Science Foundation’s Directorates of CISE, MPS, EHR, and SBE, with participation from the National Academies of Science. It is also sponsored by the South Big Data Innovation Hub, and Georgia Tech’s Institute for Data Engineering and Science.
Resources:
Previous events in the Keeping Data Science Broad series include:
View recordings of these events on the South Hub website.
Engage in upcoming 2018 South Big Data Hub activities. Join the Education Working Group or participate in a South Big Data Hub Data Carpentry Event. Increase your impact by attending a “train the trainer” session, or learn about the Hub’s recently completed installation and demo of the JupyterHub executable textbooks in support of data science education. Contact Renata Rawlings-Goss of the South Big Data Hub at rrawlings.goss@gatech.edu.
The Institute for Data Engineering and Science at Georgia Tech unifies data science researchers and resources spanning all disciplines throughout Georgia Tech to take on grand challenges in data science. It strategically builds collaborations and supporting resources to stimulate foundational research in areas such as machine learning, high-performance computing, and algorithms and optimization. It identifies and unites researchers to pursue collaborative and ambitious funding opportunities, to drive research, and to evolve and promote data science education. IDEaS provides an accessible and stable means of navigating the vast landscape of data science research and opportunities internally, and externally as it connects to industry and other partners.
The South Big Data Hub is part of a network of four regional Big Data Hubs, launched by the National Science Foundation and funded in part by host universities and other partners. Managed jointly by co-executive directors at the Georgia Institute of Technology and the University of North Carolina at Chapel Hill, the South Hub serves 16 states and the District of Columbia—from Texas to Delaware—with more than 800 members from universities, corporations, foundations, and cities committing their support. The Hubs accelerate partnerships, grow R&D communities, facilitate resource and data sharing, and build data science capacity for education and workforce development.
]]>Along with networked data transfer, memory-centric architectures, improving energy efficiency, and the path to exascale computing, one of the key topics of discussion this year is diversifying the HPC workforce.
Lorna Rivera is a research scientist at the Georgia Tech Center for Education Integrating Science, Mathematics and Computing (CEISMC) and part of the Georgia Tech contingency at SC17 this week.
Because her work brings together scientific content, teaching, and equity, Rivera has been invited to participate in a number of panels and workshops on expanding diversity in the HPC community.
In a Q&A with the College of Computing just prior to SC17, Rivera provides advice for HPC leaders and explains why reaching diverse groups requires diverse strategies.
What are some tips for increasing the number of female/underrepresented minority (URM) applicants for HPC positions?
First, recruitment is key. You have to start early, not when a position is opening soon. Be sure to review all recruitment materials for inclusive language. There are lots of resources for how to do this. Some of my favorites are from NCWIT, as well as from the WISELI group.
Some other tactics that have been successful are providing opt-in training on unconscious bias for search committees, and offering incentives to employees for recommending or recruiting diverse candidates.
Once identified, pursue all qualified candidates (underrepresented or not) equally without making potentially detrimental assumptions about their interest level in your institution/organization because of their background.
A really useful idea is having candidates meet privately with an advocate that has no bearing on the search committee decision and that complies with Equal Employment Opportunity rules. This person should go over the organization’s policies and practices without inquiring about the applicant’s family/marital status etc. and answer any work-life related questions. This exchange must be entirely confidential and inaccessible to the search committee.
What are some challenges unique to reaching underrepresented applicants?
What’s unique about reaching female and URM applicants is that the challenges are often within the search committee, not coming from the applicants. This is largely due to a lack of sensitivity and awareness, as well as bias against various URM groups. It’s important to rethink criteria like “qualified” or “excellent.” These criteria should be objective and defined prior to the search. It’s when things are unclear from the beginning that we lean toward biased interpretations of what is “qualified” and what represents “excellence”, like an Ivy League graduate or someone who resembles the status quo in visual appearance, demographics, etc.
What are some emerging strategies and/or best practices for promoting diversity and growing diverse participation in HPC?
HPC lacks reliable population data due to its interdisciplinary nature. Unlike traditional science disciplines, it’s hard to put a label on and identify someone in HPC. Right now, lots of the focus within the broader community is on defining and measuring the population in order to generate a baseline for future study. While I too am interested in this and have conducted this type of evaluation for groups like SC, my work extends to identify those that are least well served and ask why in order to improve diversity and equity within a particular context.
Is there anything that Georgia Tech is doing that can be replicated within other institutions?
CEISMC Executive Director Lizanne DeStefano and I work together on many HPC evaluation projects. The findings from these evaluations are being utilized across HPC both domestically and internationally. Our work is quite varied from reducing selection bias in competitive HPC student programs to conducting organizational climate studies of large- and small-scale HPC projects to help improve retention of URM staff/faculty.
What advice do you have for HPC leaders working to build diverse teams?
If possible, partner with or hire someone with demonstrated success in the area to help implement strategies for your group/organization. Culture change takes commitment and is very difficult to achieve doing it as a side job. Proposal requests (RFPs) often include a broadening participation component, but many projects are not adequately prepared. The HPC world is lucky to have a resource like XSEDE that has a strong broadening participation program designed to serve the entire community through partnerships and direct programming.
Despite facing the greatest obstacles and receiving the least support of underrepresented groups, women of color are often overlooked when discussing diversification. How can companies fight against the current trend of taking a one-size-fits-all approach to advancing women and women of color?
As a Puerto Rican woman, this really resonates with me. There is definitely no good one-size-fits-all approach, and companies are likely to miss out on the benefits of diversity if they fail to acknowledge that reaching diverse groups requires diverse strategies.
Developing a strategic plan to address key target populations in a stepped approach is one way of tackling this problem. Groups should try to pilot evidence-based programs and activities with each target population in isolation, and then adapt their programs for other groups. All activities should be closely monitored in order to be responsive to any successes or slippage that may occur.
Documenting these outcomes for the community is also incredibly helpful. We are in the middle of doing this now with an international HPC student-training program. The planning committee wanted to increase the number of female participants in the competitive program, but they were not being rated highly. Our investigation found gender-based selection bias and we made recommendations for reducing it, including an overhaul of the application form in 2016. That year no evidence of gender-based selection bias was found.
Workplace diversity efforts are ongoing, but progress remains slow. How can we increase the urgency and speed up the pace of these initiatives?
Organizations like Georgia Tech are setting a good example by modeling and piloting initiatives so that others can adapt based on Tech's lessons learned and hopefully reap substantial benefits. I personally would not be able to attend SC this year if it weren’t for the childcare grant offered by the College of Sciences. Because I’m able to attend, I will be giving more than 10 presentations this week (some closed/proprietary) on evaluation findings in HPC projects – all of which have an equity and diversity component. Hopefully, this work will add a few drops in the bucket to addressing this massive challenge of moving forward on these much-needed programs.
]]>404-894-7253
]]>The software, which is open-source and free to use, could be used by chemists and materials scientists to adapt algorithms and equations to run on quantum computers, which could in theory be orders of magnitude more powerful than conventional supercomputers.
Called OpenFermion, the new software contains a library of algorithms for simulating how electrons interact, which is important for work in both chemistry and materials science, on a quantum computer. Until now, these interactions could only be simulated on powerful conventional computers. Chemists would have to team up with specialized quantum developers and do a lot of coding to be able to run the equations on a quantum machine. But Google and the other developers are releasing two plug-ins that allow OpenFermion to directly translate algorithms from two of the most popular conventional simulators, Psi4 and PySCF, to run on a quantum computer.
Psi4 is an open-source quantum chemistry code developed by a multi-institution team led by Dr. David Sherrill at Georgia Tech. Dr. Sherrill worked with the Google consortium, which includes ETH Zurich, Lawrence Berkeley National Laboratory, IBM, and others this summer to assist them in utilizing Psi4 with the OpenFermion project. "We are very happy that our cooperation with them is paying off," said Sherrill. The project promises to markedly accelerate research by putting important tools within the reach of more researchers without specialized coding skills or the resources to undertake time-consuming rewriting of essential algorithms.
Read the full article at Bloomberg Technology.
]]>]]>
jsalazar@gatech.edu
]]>Gravitational waves were detected by the LIGO Scientific Collaboration (LSC), which was founded by Barish. The LSC is a group of more than 1,000 scientists around the world that includes 17 Georgia Tech faculty, postdoctoral researchers and students.
Gravitational waves are ripples in the fabric of space-time caused by cataclysmic events in the universe. Their existence was a major prediction of Albert Einstein’s 1915 general theory of relativity. Because gravitational waves carry information about their dramatic origins that cannot otherwise be obtained, their detection opened an unprecedented new window onto the cosmos.
Since 2015, LIGO has announced four gravitational wave detections, including one last week. Each was caused by violent collisions of binary black holes.
Below are reactions from Georgia Tech, including School of Physics Professor Laura Cadonati, who serves as deputy spokesperson for the LSC.
“This year’s Nobel award honors three scientists whose vision and dedication have made it possible to discover gravitational waves, explore answers to questions in fundamental physics and unlock a new window on the universe,” Cadonati says. “I am grateful for the opportunity to work with them in the realization of this vision, together with the hundreds of colleagues in the LIGO Scientific Collaboration. Today’s prize is also a celebration of collaboration and collegiality in science; we all feel honored.”
Deirdre Shoemaker
LSC member and director of Georgia Tech’s Center for Relativistic Astrophysics
“The Nobel Committee honored three people who made astronomy with gravitational waves possible,” Shoemaker says. “They laid down the theoretical foundations of gravitational-wave sources and paved the way in the design of the instruments to detect these sources. They also motivated generations of scientists to answer the call to gravitational-wave astronomy and provided the political courage to undertake the task. It is a great privilege to join in honoring their work and that of all of my colleagues in gravitational-wave physics.”
Pablo Laguna
LSC member and chair of the School of Physics
“This year’s Physics Nobel Prize recognizes three scientists that led in establishing the foundations for the monumental achievement of building an instrument (LIGO) to learn about the cosmos from observations of changes in the fabric of space-time (gravitational waves)."
Paul M. Goldbart
Dean of the College of Sciences, Betsy Middleton and John Clark Sutherland Chair and professor in the School of Physics Professor
“It is a joy to add my acknowledgement of these pioneers of gravitational radiation. Their tenacity has paved the way for our colleagues to participate in this extraordinary scientific adventure that heralds the dawn of a new era in astronomy.”
]]>Led by the Georgia Institute of Technology, the NSF Engineering Research Center for Cell Manufacturing Technologies (CMaT) could help revolutionize the treatment of cancer, heart disease, autoimmune diseases and other disorders by enabling broad use of potentially curative therapies that utilize living cells – such as immune cells and stem cells – as “drugs.” Examples of these highly promising therapies include T cell-based immunotherapies for blood cancers, such as the one developed at the University of Pennsylvania and approved in August by the U.S. Food & Drug Administration, and a gene-modified stem cell therapy recently approved in Europe for a form of the so-called “bubble boy” syndrome.
To facilitate the widespread application of these cutting-edge emerging treatments, CMaT will develop robust and scalable technologies, innovative analytical tools, and engineering systems that will enable industry and clinical facilities to reproducibly manufacture efficient, safe and affordable cell-therapy products. The center, one of four ERCs announced September 12 by the NSF, will also develop improved models for a robust supply chain, storage and distribution system for these therapeutic cell products.
“For over 30 years, NSF Engineering Research Centers have promoted innovation, helped to maintain our competitive edge, and added billions of dollars to the U.S. economy,” said NSF Director France Córdova. “They bring together talented innovators and entrepreneurs with resources from academia, industry and government to produce engineers and engineering systems that solve real-world problems. I am confident that these new ERCs will strengthen U.S. competitiveness for the next generation and continue our legacy of improving the quality of life for all Americans.”
In addition to the consistent manufacture of cell-based therapies, the public-private CMaT initiative will also help develop a skilled, diverse and inclusive bio-manufacturing workforce through extensive education and training activities at the K-12, technical college, undergraduate, graduate and postdoctoral levels.
Living cells become “drugs”
“Unlike pharmaceuticals and other products now used in medical treatments, cells are living entities whose properties can significantly change depending on nuances in the way they are grown, stored or otherwise manipulated,” said Krishnendu Roy, director of CMaT and the Robert A. Milton chair professor in the Wallace H. Coulter Department of Biomedical Engineering at Georgia Tech and Emory University. “The center will develop new engineering tools and scalable methods to better characterize, expand, differentiate, separate, transport and store high-quality cells so they provide consistent therapeutic effects, allowing them to be used in standardized therapies by clinicians to serve large numbers of patients worldwide.”
Beyond Georgia Tech, the center will include major partners – the University of Georgia, the University of Wisconsin-Madison and the University of Puerto Rico, Mayaguez Campus – as well as affiliate partners such as the University of Pennsylvania, Emory University, the Gladstone Institutes and Michigan Technological University. Additional international academic partners, as well as industry and the U.S. national laboratories, will also be critical collaborators in the effort.
Moving discoveries into application
“Georgia Tech has a long history of building collaborative partnerships with industry, the national labs and other research universities. With the support of the NSF and this new ERC, we will be able to capitalize on expertise in multiple areas, taking transformative research from the laboratory to practice much more quickly,” said Georgia Tech President G. P. “Bud” Peterson. “The Center for Cell Manufacturing Technologies will also help us educate, train and prepare the workforce in a new industry, thereby continuing to strengthen the U.S. economy.”
Clinical trials have already established the effectiveness of several cell-based therapies and many other trials are underway. But for these exciting therapies to advance into broad healthcare use, the cells will have to be produced in much larger quantities and with more consistent quality than is now available. There are also very few, if any, established industry standards for analytics and processes in cell manufacturing, which hinders consistent production of safe and efficacious cells. Another key limitation identified by industry is the need for a highly-trained workforce.
CMaT would address these barriers through transformative innovations that build upon a series of earlier efforts, including the National Cell Manufacturing Consortium (NCMC) roadmap, infrastructure established at Georgia Tech with support from the Marcus Foundation, quality and other standards programs from the National Institute of Standards and Technology (NIST) and independent industry-led bodies, and translational activities by industry, entrepreneurs and other partners.
The NSF’s multidisciplinary engineering research centers address unique, complex engineering challenges by stimulating knowledge and tech transfer between different sectors, from electronics to energy to infrastructure. Each center takes on a specific engineering research challenge.
“The overall goal of the NSF Engineering Research Centers program is nothing less than to revolutionize engineering research and education in the United States,” said Dawn Tilbury, NSF assistant director for engineering. “We look forward to the exciting advances and outcomes in these important areas.”
Accelerating clinical trials
Beyond established cell-based therapies, the work of CMaT should accelerate the development of new therapies and the testing needed to bring them into the clinic, said Steven Stice, director of the University of Georgia’s Regenerative Bioscience Center (RBC). Regenerative medicine applications could offer new ways of treating diseases for which there are now essentially no treatments, including Parkinson's, Alzheimer’s, heart disease and stroke.
“There are a significant number of cell therapy clinical trials and investments in the field,” Stice said. “But there is little or no investment in a set of consistent standardization methods to optimize how these therapies should work. For instance, we know that cell therapies will improve human health, but right now it’s difficult to guarantee that each dose produced will be as potent as the next. The work done by CMaT researchers will help solve some of these problems.”
The University of Pennsylvania develops cellular therapies and has conducted more than 40 clinical trials of cell-based therapies, including those for engineered T cell therapies and chimeric antigen receptor (CAR) T cells. An example is recently-approved treatment for relapsed and refractory acute lymphoblastic leukemia in pediatric and young adult patients.
“The cell and gene therapy fields are on the cusp of multiple regulatory approvals in the near term,” said Bruce Levine, Barbara and Edward Netter Professor in Cancer Gene Therapy in the Perelman School of Medicine at the University of Pennsylvania. “The challenges ahead lie in developing manufacturing and testing processes incorporating automation that can bring costs down and allow access to more patients.”
Developing broad-based innovations
Critical innovations often occur at the boundaries of disciplines, and CMaT will bring together relevant specialties for both research and workforce development, noted Madeline Torres-Lugo, a professor in the Department of Chemical Engineering at the University of Puerto Rico, Mayaguez Campus.
“Due to the complexity of cells as living organisms, a team with a strong background in biology, chemistry, physics, materials science, and engineering is required for this initiative,” Torres-Lugo said. “Our participation and contribution to CMaT will ensure that Puerto Rico not only remains at the forefront of pharma manufacturing, but also supports cell manufacturing technologies here and around the world by educating highly talented engineering students.”
CMaT testbeds have been selected to address several cell types that are in early stages of clinical adoption or moving toward clinical applications, but it isn't yet clear what cell types will have the greatest therapeutic impacts, noted Sean Palecek, the Milton J. and A. Maude Shoemaker Professor in chemical and biological engineering at the University of Wisconsin-Madison. Therefore, one of the center’s challenges will be to ensure that fundamental discoveries, and tool and technology development efforts, will apply to multiple cell types.
“Our work will provide safer and more potent cell products that will allow clinical studies to establish the effectiveness of these cells as therapeutics,” Palecek said. “In addition, our work on scaling cell production will enable manufacturing of sufficient numbers of cells to replace damaged organs, such as the loss of heart muscle after a heart attack, at a cost that makes these therapies accessible to broad segments of society. We will also train the future leaders of the emerging therapeutic cell manufacturing industry. These students and their work establishing this industry will be the most significant impact of CMaT.”
New centers among 19 ERCs
Since the program’s inception in 1985, NSF has funded a total of 74 ERCs and will support 19 in this fiscal year, including four new centers. Each center receives NSF funding for up to 10 years. During this time, centers build partnerships with industry, universities and other government agencies that will sustain them for years to come.
In May, the National Academies published a report, “A new vision for center-based engineering research,” which was the result of an NSF-funded study to examine the future of the NSF ERC program.
The report identifies and recommends strategies to enable NSF multidisciplinary engineering research centers to continue addressing key research, education and innovation needs of the United States in a changing global context.
“ERCs are widely known as outstanding examples of successful partnerships between universities, private industry and government that have made significant contributions to address national challenges,” said Don Millard, acting division director for the NSF Division of Engineering Education and Centers. “We are continually working with the scientific and engineering communities, as well as private industry and government partners, to ensure NSF-funded centers and grantees are best-equipped to match societal needs with research abilities.”
Research News
Georgia Institute of Technology
177 North Avenue
Atlanta, Georgia 30332-0181 USA
Media Relations Contacts: John Toon (404-894-6986) (jtoon@gatech.edu) or Ben Brumfield (404-660-1408) (ben.brumfield@comm.gatech.edu).
Writer: John Toon
]]>Research News
(404) 894-6986
]]>In partnership with the Institute for People and Technology (IPaT), Georgia Tech’s initiative on Smart Cities and Inclusive Innovation is developing innovative approaches to shaping resilient and sustainable communities.
Through research and development, strategic partnerships, and cutting-edge programming, the initiative brings Georgia Tech's interdisciplinary expertise in technology and policy to the development of smart cities and communities.
The Institute has assembled a 20-plus member interdisciplinary faculty council co-chaired by Gisele Bennett, Georgia Tech’s associate vice president for research faculty interaction, and Beth Mynatt, executive director of IPaT.
Members of the council include the College of Design’s Ellen Dunham-Jones, Catherine Ross, and Dennis Shelden, as well as representatives from four Georgia Tech colleges and nine schools.
Georgia Tech is pairing the council with a number of key partnerships, including the City of Atlanta, which recently launched the North Avenue Smart Corridor pilot project. As the City’s official research partner on the project, Georgia Tech is helping to develop, deploy and evaluate smart technologies aimed at improving public safety, environmental health and traffic congestion along the corridor.
As part of the initiative, a new Smart Cities and Inclusive Innovation website provides a comprehensive summary of key partnerships, projects, and data sets. Visit smartcities.gatech.edu for more information.
]]>The Center for Research into Novel Computing Hierarchies (CRNCH), led by Computer Science Professor Tom Conte, is spearheading the Rogues Gallery as one of its first initiatives since being founded in November 2016. It will be a collection of some of the most unique computers in the field. These machines are so rare, only a few know how to program them, or they are so new, no one has programmed them. The goal is simple: to collect unusual hardware to make it accessible to the industry and academia.
With an Intelligence Advanced Research Projects Activity (IARPA) grant of $662,525, CRNCH researchers can finally purchase their first rogue, the Emu Chick.
This is a memory-centric architecture that employs threads — not processors — to move massive irregular data sets, effectively expediting data analysis workloads. The Chick could eventually be used to tackle fraud detection, genome sequencing for personalized medicine, real-time portfolio valuation and trading, and software development for larger Emu systems. But the machine is so new that its full capabilities aren’t known yet, making it an ideal Rogues Gallery resident.
Computational Science and Engineering senior research scientist Jason Riedy hopes to study the Chick’s performance, scalability, and programmablity with massive graphs and multilinear modeling for data analysis. He won’t be the only one testing the Chick’s capabilities.
The Rogues Gallery is designed to be open to all researchers, who could access the machine remotely if needed. Students will also have the chance to work on some of the most cutting edge machines, making them more competitive in the job market.
“The goal is to maintain Georgia Tech’s presence in novel and unusual architecture,” Riedy said.
CS research scientist Jeffrey Young and researchers at Georgia Tech Research Institute are working to set up a small cluster of Field Programmable Gate Array (FPGA) devices with 3-D stacked memory. These devices allow for programming custom, reconfigurable hardware, so that sparse linear solvers and graph analytic algorithms can be run at a faster rate. The team expects to have the Rogues Gallery open by late fall, and is exploring opportunities for other novel hardware related to embedded systems, neuromorphic, and quantum computing.
“When you have the hardware and expertise, people come to you,” Young said.
The need to efficiently analyze large, complex data sets is a growing challenge, and one that scholars from all areas need to tackle. With this in mind, Georgia Tech launched the Transdisciplinary Research Institute for Advancing Data Science (TRIAD) on Oct. 3.
TRIAD is part of a larger National Science Foundation (NSF) project, the Transdisciplinary Research in Principles of Data Science, in which 14 institutions across 11 states collaborate on long-term big data research and training. With a $1.5 million NSF award, Georgia Tech is able to bring researchers from four of its six colleges together, representing fields such as theoretical computer science, mathematics, and statistics.
“This diversity is one of the key elements of this project,” said Stephen Cross, executive vice president for research. “New ideas come from being at the boundaries of all these fields.”
The College of Computing (CoC) has been involved in the project since the beginning. School of Computational Science and Engineering (CSE) Professor Srinivas Aluru and School of Computer Science (SCS) ADVANCE Professor of Computing Dana Randall are on TRIAD’s management team. As co-directors of the Institute for Data Engineering and Science, which submitted the NSF proposal, they already research the impact of big data but praise TRIAD’s interdisciplinary approach.
“Everyone on campus is doing something involving algorithms,” Randall said. “You can use an existing algorithm, but if you think about a problem from scratch, you can make breakthroughs.”
The community research aspect of TRIAD will be carried out by up to 40 professors and more than 250 students. The research will focus on four issues:
The first phase of TRIAD will facilitate interdisciplinary research through working groups; national and international week-long workshops featuring research presentations, tutorials, poster sessions, and panels; and organized innovation labs.
SCS Assistant Professor Jake Abernethy, School of Industrial and Systems Engineering Professor Sebastian Pokutta, and Mathematics Professor Prasad Tetali (who also holds a joint appointment in SCS) will host the first workshop on machine learning, optimization, and decision making from March 5 to 9, 2018.
Also look forward to more CoC workshops on campus throughout the spring, such as the one on algorithms and randomness led by SCS faculty Algorithms & Randomness Center Director Eric Vigoda, Professor Santosh Vempala, and Tetali from May 14 to 17, 2018.
Other CSE faculty will serve as senior investigators: Bistra Dilkina and Haesun Park.
“Our goal is to have Georgia Tech be seen as the go-to place for data science and machine learning,” said Professor Irfan Essa, associate dean for research in the College of Computing and director of the Center for Machine Learning at Georgia Tech. “This is something we’re really engaged with and invested in.”
For the 2017-18 academic year, GT Computing is welcoming 10 new tenure-track faculty members. With specialties ranging from machine learning, algorithms, and high-performance computing to artificial intelligence, robotics, and cryptography, these new faculty members bring a diverse set of interests and experience to the GT Computing community.
“Our new faculty members are among the best of the best in their respective fields and their collective teaching abilities and research acumen are highly sought after from industry and academia,” said John P. Imlay Jr. Dean of Computing Zvi Galil. “Each of them is certain to be a true asset to the College of Computing community and we are delighted to welcome them to Georgia Tech.”
Of the 10 new tenure-track faculty members, two are in the School of Computational Science and Engineering (CSE), five are from School of Computer Science (CS), and three in the School of Interactive Computing (IC).
New CSE faculty members include Assistant Professor Tobin Issac and Professor Felix Herrmann. Hermann holds a joint appointment between CSE and the schools of Earth and Atmospheric Sciences and Electrical and Computer Engineering.
More information about CSE’s new faculty members is available.
Assistant Professors Jacob Abernathy, Xu Chu, and Jamie Morgenstern, and Associate Professor Vladimir Kolesnikov have joined the CS faculty ranks. And, Professor Vivek Sarkar has taken a position in CS as a Stephen Fleming Chair for Telecommunications.
More information about CS’s new faculty members is available.
The new School of IC faculty members are Assistant Professors Sauvik Das and Matthew Gombolay. Professor Seth Hutchinson now fills the KUKA Chair for Robotics in the School of IC and is the associate director of the Institute for Robotics and Intelligent Machines.
More information about IC’s new faculty members is available.
]]>404-894-7253
]]>Efforts to launch the theory-focused institute were supported by the Institute for Data Engineering and Science (IDEaS) with faculty from the Colleges of Engineering, Sciences and Computing. Xiaoming Huo, principal investigator and professor in the H. Milton Stewart School of Industrial & Systems Engineering (ISyE), will serve as the executive director of TRIAD. “The establishment of TRIAD is tremendously beneficial considering data science is a phenomenon that brings with it so many opportunities. There is a lot of research that needs to be conducted in this emerging field, and we will focus on building the theoretical foundations to establish the principles of data science,” said Huo.
“With the launch of TRIAD and the already well-established NSF South Big Data Innovation Hub, Georgia Tech plays a key role nationally in data science from theory to applications to building community partnerships,” said Srinivas Aluru, co-executive director of IDEaS and a professor in the School of Computational Science and Engineering. Dana Randall, fellow co-executive director and professor in the School of Computer Science said, “Advances in data science foundations can lead to revolutionary new solutions for analyzing data. TRIAD will accelerate this research by bridging fields to expand our collective capabilities.”
Both Aluru and Randall serve as co-principal investigators of TRIAD, along with Prasad Tetali, a professor in the School of Mathematics and School of Computer Science, and Jeff Wu, the Coca-Cola Chair in Engineering Statistics and a professor in ISyE. “The emphasis on theoretical foundations (of data science) offers a great opportunity for mathematicians to actively engage with other scientists and help make breakthroughs in this fast growing interdisciplinary field," said Tetali.
The analysis of massive, dynamic, noisy and complex data arising in virtually every sphere of human activity is a pressing problem of our time. NSF is responding by dedicating $17.7 million in funding for 12 Transdisciplinary Research in Principles of Data Science (TRIPODS) institutes, including TRIAD. Encompassing 14 institutions in 11 states, these projects will promote long-term research and training activities in data science that transcend disciplinary boundaries.
“Data is accelerating the pace of scientific discovery and innovation,” said Jim Kurose, NSF assistant director for Computer and Information Science and Engineering (CISE). “These new TRIPODS projects will help build the theoretical foundations of data science that will enable continued data-driven discovery and breakthroughs across all fields of science and engineering.”
According to NSF, the TRIPODS awards will enable data-driven discovery through major investments in state-of-the-art mathematical and statistical tools, better data mining and machine learning approaches, enhanced visualization capabilities and more. They also support innovative educational pathways to train the next generation of data scientists.
TRIAD will bring together senior, mid-career and junior faculty members, postdoctoral fellows, graduate and undergraduate students and data science practitioners at large using focused working groups, national and international workshops and organized innovation labs. It will build an intellectual atmosphere to connect stakeholders from across the nation and the world on a regular basis.
Initially, TRIAD will focus on four research topics: advanced mathematical modeling for contemporary data; new inferential strategies that can be both scalable and de-centralized; efficient optimization tools with theoretical guarantees; and applications in the context of large datasets from domains including biology, design, manufacturing, logistics and sustainability.
TRIAD will launch with 39 committed faculty members from across Georgia Tech, including from ISyE, the School of Electrical and Computer Engineering, the George W. Woodruff School of Mechanical Engineering, the School of Mathematics, the School of Biological Sciences, the School of Computational Science and Engineering and the School of Computer Science.
NSF’s award establishes TRIAD as a Phase I investment. It may be subsequently expanded to Phase II through a second competitive proposal process.
]]>Director of Communications and Grant Writing, Institute for Data Engineering and Science
]]>As in years past, the College of Computing had the opportunity to share some of its recent achievements with President Peterson and we thought you might like to see them, as well.
The slides below provide only a small peek at the dynamic people, innovative research, and worthwhile results at the GT Computing, but you can keep up with all of this and more by following the College on Facebook, Instagram, and Twitter.
The full script and slide presentation for the 2017 Institute Address are available on the president's website.
]]>404-894-7253
]]>]]>
IDEaS Director of Communications and Grant Writing
]]>IDEaS Director of Communications and Grant Writing
]]>The event was set in motion by Executive Vice President of Research Steve Cross, whose welcome included a view of interdisciplinary research across Georgia Tech, and the rationale and value of the Interdisciplary Research Institutes (IRIs) like IDEaS. A series of short introductory presentations transitioned into a town hall conversation about opportunities, resources, and ways to engage. The evening ended with a celebratory reception and presentation of data-driven music provided by Mason Bretan, a Ph.D. student in the Center for Music Technology. The reception also included a presentation featuring the research of IDEaS-affilated faculty, centers, and institutes.
During the town hall, Randall introduced the IDEaS team, including the recently appointed associate directors David Sherrill (Chemistry and Biochemistry), Deirdre Shoemaker (Physics), and Marilyn Wolf (Electrical and Computer Engineering), and the newly formed IDEaS Faculty Council. Randall also talked about the role IDEaS will play in shaping research neighborhoods in the upcoming Coda Building.
"We want the community to engage with us and start a conversation about their data science needs--to create the next big idea in data science with us," Randall said.
Aluru gave an overview on how the Institute is composed, and of its motivation, goals, and affiliations with centers dedicated to data science foundational research and applications. IDEaS is also closely allied with the South Big Data Hub, with Aluru (as the hub PI) and Renata Rawlings-Goss (as the hub co-executive director) serving in leadership roles for both entitites.
Aluru discussed IDEaS's progress to date, including three industry-academia-bridging workshops in the areas of materials and manufacturing, analytics and machine learning, and precision medicine. A fourth upcoming workshop at Georgia Tech on April 26-27 will concentrate on IoT for smart and connected cities and campuses.
Aluru summarized IDEaS's work developing collaborative proposals to date, and issued a call to mobilize for future strategic activities. Aluru and Randall called attention to upcoming activities such as several IDEaS mini-retreats: IDEaS in Research, IDEaS for the Future, a retreat on Data Science for Pediatrics (with the Pediatric Technology Center) on May 23, and a Retreat for Energy (with SEI) in early Fall. Other events include monthly networking gatherings, with the first of these scheduled for April 28 at 4 p.m.; Data Science modules, short courses, specialized training; and the IDEaS seminar series starting in the fall on Fridays from 3 p.m. to 4 p.m. IDEaS is also planning to launch the IDEaS Labs, and the IDEaS Data Repositories. Anyone interested in contributing to these efforts can contact the IDEaS team.
Rawlings-Goss, directory of industry partnerships, gave a synopsis of the IDEaS Industry Alliance Program. "We've had some early successes," she said. "For example, through our new Program to Empower Partnerships with Industry (PEPI) program, companies like United Healthcare and McKesson are paired with faculty for research, going on-site to target relevant problems and growing new working partnerships."
She also explained how IDEaS worked with Microsoft to become an executive partner and provided faculty resources as well as cloud support. Microsoft's gift to Georgia Tech included Microsoft Azure and funding for seed projects in Data Engineering and Science. Rawlings-Goss asked interested industry and faculty researchers to get involved by contacting her.
Nearly 130 people attended the days events, with about a quarter of those from industry.The event was planned by Shkina Halbert of IDEaS, with assistance from the College of Computing Event Coordinator Birney Robert. Robert, who is also a studio artist, crafted technology-themed artwork using an arrangement of surplus store circuit boards, wiring, and light bulbs, and used these as table centerpieces.
IDEaS unifies data science researchers and resources spanning all disciplines throughout Georgia Tech to take on grand challenges in data science. It strategically builds collaborations and supporting resources to stimulate foundational research in areas such as machine learning, high-performance computing, and algorithms and optimization. It identifies and unites researchers to pursue collaborative and ambitious funding opportunities, to drive research, and to evolve and promote data science education. IDEaS provides an accessible and stable means of navigating the vast landscape of data science research and opportunities internally, and externally as it connects to industry and other partners.
While the IRI was officially organized in summer 2016, the March 29 event marked the operational launch of IDEaS and the beginning of a new campus-wide conversation on data science research, partnerships, resources, and education.
The 12 Interdisciplinary Research Institutes (IRIs) of Georgia Tech are responsible for bringing together a mix of researchers – spanning colleges, departments, and individual labs – around a single core research area. IRIs also connect a large portfolio of basic and applied research programs, support world-class research facilities and laboratories, engage Georgia Tech students, and collaborate with government and industry research partners.
Click here to view a photo album from the IDEaS launch!
]]>IDEaS Director of Communications and Grant Writing
]]>“This fourth summer of the program has been a huge success,” said Ellen Zegura, a professor in the College of Computing, and the one who began Atlanta’s DSSG program. She kicked off the final event, which shared innovative data-driven approaches and results for five projects in the areas of housing justice, food security, crowdsourced environmental monitoring, flood prediction, and building energy consumption. Video interviews of participants and an image gallery of the event are available.
The Atlanta Data Science for Social Good program is an intensive, ten-week paid internship experience that blends data science and technology design. Students are placed on multi-disciplinary teams and matched with a supervising professor to address real-world problems with partners in the City of Atlanta and local non-profit organizations. DSSG student teams blend expertise on technical topics, public policy, social issues, and education for a truly interdisciplinary approach. Bistra Dilkina, a DSSG co-director said “DSSG connects the classroom with real problems of deep community relevance. We hope this will inspire them to pursue their technical education further, and to be engaged global citizens that use their education for societal impact.”
The program addresses several emerging concerns simultaneously: finding innovative solutions that serve immediate needs, getting the experience necessary for developing the nation’s data science workforce, and helping students communicate effectively by working with actual clients and team members. “It is a unique opportunity to make a real impact in the Atlanta Community. Students get the ability to understand how to communicate data problems with organizations that are not coming at this with a high level of technical background,” said Co-Director Chris LeDantec.
The highly competitive internship selected 17 interns from around the country from a pool of over a hundred applicants with backgrounds in computer science, statistics, digital media, public policy, civil engineering, industrial engineering and urban planning. This year, DSSG received sponsorship from the National Science Foundation and the South Big Data Innovation Hub at the program level.
During the first week, students participated in hands-on tutorials in data analytics, web stack development, spatial information, and geographic information systems. In the remaining weeks, students worked with government and nonprofit partners to design data-driven solutions. They also attended weekly seminars with invited speakers from academia and industry on technical topics and prominent examples of using data for good.
By working with real data and organizations, the students develop useful tools. “Some of these apps will be used by the actual agencies, so that means they’re really influencing people’s lives,” said South Big Data Hub Co-Executive Director Renata Rawlings-Goss.
They also learn critical skills such as stakeholder engagement, data acquisition and processing, data analysis and visualization, machine learning for predictive modeling, writing, and communicating results to nontechnical audiences.
During the summer of 2017, students worked on projects targeting five social problems:
Food for Thought: Analyzing Public Opinion on the Supplemental Nutrition Assistance Program
Interns: Miriam Chappelka, Jihwan Oh, Dorris Scott, Mizzani Walker-Holmes
Georgia Tech Program Mentor: Carl DiSalvo, Associate Professor in the Ivan Allen College of Liberal Arts
Community Partner: Atlanta Community Food Bank
A major tool in the fight against hunger is the Supplemental Nutrition Assistance Program (SNAP), commonly known as food stamps. Understanding the news cycles, geography and changing attitudes of the population toward SNAP is useful for shaping policy. Teams used data from census, voting records, news articles, Twitter and Facebook to categorize sentiments and conduct text mining. The team created three products: an InfoMap to view the geography and some correlates of SNAP-related attitudes, a voting records database to prepare for meetings with policymakers, and a Visualization of Sentiment Analysis to assess sentiment about SNAP during major events.
Cycle Atlanta: Seeing Like a Bike
Interns: Javier Argota, Myeong Lee, Noel Mannariat, Erica Pantoja
Georgia Tech Program Mentor: Chris LeDantec, Associate Professor in the Ivan Allen College of Liberal Arts
Community Partner: City of Atlanta
Not all bikers feel comfortable cycling throughout Atlanta, especially during rush hour. The team designed multi-sensor boxes to generate data on factors such as traffic speed, traffic volume, percent of heavy vehicles, proximity to traffic, lane configurations, pavement conditions, and many others that account for riders’ stress. Their analyses involved the creation of a Level of Traffic Stress (LTS) model and the use of two machine learning algorithms. The goal is to provide reliable data-driven evidence that helps policy makers refine cycling infrastructure and environmental conditions for cyclists of all comfort levels.
Atlanta Housing Justice: The Anti-Displacement Tax Fund
Interns: Jeremy Auerbach, Hayley Barton, Takeria Blunt, Vishwamitra Chaganti, Bhavya Ghai
Georgia Tech Program Mentors: Christopher Blackburn, Ph.D. Candidate in the Ivan Allen College of Liberal Arts; Amanda Meng, Research Associate in the College of Computing; Ellen Zegura, Professor in the College of Computing
Community Partner: Atlanta Legal Aid Society
Urban revitalization on Atlanta’s Westside, including construction of the Beltline and a new stadium, has caused increased property values and fears that residents will be displaced. An anti-displacement tax fund exists to offset the increase in property taxes for eligible residents. However, stakeholders need to understand program cost, and number of people impacted. This project determined the number of eligible homeowners using location, owner-occupancy, property liens data, and an income model based on home characteristics. It forecasted home appreciation and property tax increases, and created an interactive web app for community members to view eligibility and estimated property tax increases. The team’s projections show a higher cost and greater number of eligible households than those previously released, highlighting the value of data science techniques and community participation.
Building Energy Analytics
Interns: Keyan Halperin, Lingzi Hong, Brendon Machado, and Ricardo Macias
Georgia Tech Program Mentor: Bistra Dilkina, Assistant Professor in the College of Computing
Community Partner: Georgia Tech Facilities Management
All Georgia Tech buildings have sensors that assess how much energy is being used every 15 minutes over the past few years. Despite the ongoing data collection, it has largely remained unexamined. The team used the energy data from four buildings, along with local weather and building occupancy to model energy usage at Georgia Tech. They determined factors related to energy usage and predicted how it changes with related events. Using six models, they predicted energy usage, common denominators, and unique aspects of a particular building that accounted for the most energy consumption. For example, the number of classes was by far the most important predictor of energy consumption in all buildings. But temperature impacted the buildings differently. Future work can extend models throughout campus, and identify inefficient buildings for upgrades.
Predicting and Alleviating Road Flooding in Senegal
Interns: Keyan Halperin, Lingzi Hong, Brendon Machado, and Ricardo Macias
Georgia Tech Program Mentor: Bistra Dilkina, Assistant Professor in the College of Computing at Georgia Tech
Community Partner: United Nations Global Pulse
Climate change has the potential to raise the risk of flood for coastal countries, impacting the living environment and threatening the success of crucial city development. The team created models to determine which populations or regions are most vulnerable to disruptions to movement caused by flooding, and which roads should be targeted for mitigation. First they modeled the flooding risk of each road segment based on historical weather and topographical data. They then modeled the volume of traffic of each road to quantify how necessary it is for accessibility between different parts of the country. Road importance was established by phone records that included when each tower location was used by a customer, traffic flow was assigned based on the data volume. Putting it all together, the team determined the consequences and mitigation strategies for traffic flows redirected due to flooded, inaccessible roads.
Several DSSG-ATL projects from previous years have received recognition:
DSSG-ATL is part of a broader community of Data Science for Social Good programs, including the original DSSG program organized by the University of Chicago, as well as the newer programs at the University of Washington, the University of Massachusetts Amherst, and most recently at the IBM T.J. Watson Research Center. Participants become part of an exciting network of students, mentors, professors and projects taking place around the United States. DSSG-ATL maintains online descriptions of its projects on its website.
]]>
The project is one of 10 “Big Data Spokes” announced by the National Science Foundation (NSF). The NSF’s $10 million initiative was created to improve the ability to solve the nation’s most pressing challenges with the use of big data. The Georgia Tech, Smithsonian and IBM “Spoke” will receive $1 million from NSF. IBM will also provide in-kind gifts. Overall, the project engages 24 researchers from 14 institutions from academia, industry, government and non-profit organizations.
“Environmental sustainability is a growing concern for our country. Scientists and citizens needbetter tools and data to rapidly build and test conceptual models of ecological phenomena,” said Ashok Goel, a Georgia Tech professor who is the principal investigator of the collaboration. “We want to empower them.”
The Encyclopedia of Life (EOL), headquartered at the Smithsonian Institution, is an online, open-access database that gathers information about all biological species on Earth.
“Modelers tell us that predicting an ecosystem's response to global changes requires knowledge of things like the mass of an algal cell, the lifespan of a copepod and the ecological partners of a reef-building coral,” said Bob Corrigan, EOL’s director of operations. “EOL is surfacing, structuring and sharing hundreds of years of careful measurements by generations of biologists. Combining these assets with the capabilities of Georgia Tech and IBM will give scientists and students alike the ability to model and study our biosphere at scales that have not been possible before.”
As part of the Spoke project, Watson Developer Cloud’s Language and Vision services will be trained to deeply understand the specialized ecology domain represented in the EOL webpages and images.
“Unlocking all of this unstructured information from the Smithsonian’s Encyclopedia of Life, bringing it into the context of other relevant structured knowledge, and making it available for further human and machine reasoning holds tremendous potential,” said Lisa Amini, director, Cognitive Computing: Knowledge and Reasoning at IBM Research. “The possibilities are endless.”
Users will then take that information and plug it into Georgia Tech’s Modeling & Inquiry Learning Application (MILA) system. The interactive tool allows scientists to rapidly generate conceptual models, evaluates them through simulation and provides results.
The NSF grant will allow the team to seamlessly link EOL, Watson and MILA. The goal is to build a working system that enables ecological modeling by early 2018.
“You can have all the information in the world, but if you can’t easily find the knowledge, you can’t build a model,” said Goel. “And if you can’t build a good model, the information is useless. Our project uses artificial intelligence to address these concerns.”
The Big Data Spokes program is supported and organized by the NSF’s Big Data Regional Innovation Hubs (BD Hubs). The four Hubs (South, Northeast, Midwest and West) foster multi-sector collaborations among academia, industry and government. Georgia Tech co-leads the South Hub with the University of North Carolina.
"The Big Data Spokes advance the goals and regional priorities of each BD Hub, fusing the strengths of a range of institutions and investigators and applying them to problems that affect the communities and populations within their regions," said Jim Kurose, assistant director of NSF for Computer and Information Science and Engineering. “We are pleased to be making this substantial investment today to accelerate the nation’s big data R&D innovation ecosystem.”
Two other Spoke awards have ties to Georgia Tech. Santiago Grijalva, Georgia Power Distinguished Professor in the School of Electrical and Computer Engineering, will study smart grids using big data with Texas A&M. Gari Clifford, an Emory University associate professor with a joint appointment in the Wallace H. Coulter Department of Biomedical Engineering, will investigate how to use data from fitness trackers and environmental monitors to improve patient care.
“Georgia Tech’s inclusion in these awards is reflective of the Institute’s unique breadth and depth of expertise that spans all areas of data science and data-driven discovery,” said Srinivas Aluru, co-executive director of Georgia Tech’s Institute for Data Engineering and Science and principal investigator of the South Big Data Hub.
]]>