{"571701":{"#nid":"571701","#data":{"type":"event","title":"ISyE Seminar - Sasha Rakhlin","body":[{"value":"\u003Cp\u003ETITLE:\u0026nbsp; Online Prediction: A Marriage of Optimization and Probability\u003C\/p\u003E\r\n\r\n\u003Cp\u003EABSTRACT:\u003C\/p\u003E\r\n\r\n\u003Cp\u003EThe talk will focus on two seemingly unrelated topics: (i) existence\u0026nbsp;of prediction strategies that have a performance guarantee for all\u0026nbsp;sequences and (ii) probabilistic inequalities for martingales. We will\u0026nbsp;point to a certain equivalence between these two topics, with the most\u0026nbsp;basic example going back to the work of T. Cover in 1965. In light of\u0026nbsp;the equivalence, we will develop computationally efficient prediction\u0026nbsp;methods for problems with a combinatorial benchmark, even when\u0026nbsp;estimating the correct model is NP-hard. Exploiting the equivalence in\u0026nbsp;the other direction, we show that tail bounds for a certain ratio-type\u0026nbsp;inequality follow with ease from mirror descent with an adaptive step\u0026nbsp;size.\u003Cbr \/\u003E\r\n\u003Cbr \/\u003E\r\nJoint work with K. Sridharan\u003C\/p\u003E\r\n\r\n\u003Cp\u003E\u003Cstrong\u003EBio:\u0026nbsp; \u003C\/strong\u003EAlexander (Sasha) Rakhlin is an Associate Professor of Statistics\u0026nbsp;at the University of Pennsylvania, The Wharton School. He received his\u0026nbsp;Bachelors from Cornell University, a Ph.D. from MIT, and joined Penn\u0026nbsp;after working as a postdoctoral researcher at UC Berkeley. Sasha\u0026rsquo;s\u0026nbsp;interests span a range of topics, including statistics, machine\u0026nbsp;learning, online prediction, and optimization.\u003Cbr \/\u003E\r\n\u003C!--[if !supportLineBreakNewLine]--\u003E\u003Cbr \/\u003E\r\n\u003C!--[endif]--\u003E\u003C\/p\u003E\r\n","summary":null,"format":"limited_html"}],"field_subtitle":"","field_summary":"","field_summary_sentence":[{"value":"ISyE Seminar - Sasha Rakhlin"}],"uid":"27187","created_gmt":"2016-09-01 10:21:52","changed_gmt":"2017-04-13 21:14:51","author":"Anita Race","boilerplate_text":"","field_publication":"","field_article_url":"","field_event_time":{"event_time_start":"2016-11-02T16:00:00-04:00","event_time_end":"2016-11-02T16:00:00-04:00","event_time_end_last":"2016-11-02T16:00:00-04:00","gmt_time_start":"2016-11-02 20:00:00","gmt_time_end":"2016-11-02 20:00:00","gmt_time_end_last":"2016-11-02 20:00:00","rrule":null,"timezone":"America\/New_York"},"extras":[],"groups":[{"id":"1242","name":"School of Industrial and Systems Engineering (ISYE)"}],"categories":[],"keywords":[],"core_research_areas":[],"news_room_topics":[],"event_categories":[{"id":"1795","name":"Seminar\/Lecture\/Colloquium"}],"invited_audience":[{"id":"78761","name":"Faculty\/Staff"},{"id":"78751","name":"Undergraduate students"},{"id":"174045","name":"Graduate students"}],"affiliations":[],"classification":[],"areas_of_expertise":[],"news_and_recent_appearances":[],"phone":[],"contact":[{"value":"\u003Cp\u003EArkadi Nemirovski\u003C\/p\u003E\r\n","format":"limited_html"}],"email":[],"slides":[],"orientation":[],"userdata":""}}}