CIS & MINDS Seminar - Dmitriy Drusvyatskiy

Recorded seminar link: <br><a href=" Dmitriy Drusvyatskiy </p><p>       Professor</p><p>       University of Washington</p><p><span>Topic:  “Optimization for large-scale learning: beyondsmoothness and convexity”</span></p><p>Abstract: Estimation and learning algorithms aredramatically improving our capacity to extract information from massivedatasets, with impressive consequences for technology and society at large.Although these algorithms have had widespread empirical success, we have yet tofind a coherent mathematical foundation thatcan explain whythese algorithms succeed on such a wide array of problems. The challenge isthat the two assumptions that underpin classical optimizationtheory---smoothness and convexity---rarely hold in contemporary settings.Nonetheless, simple optimization algorithms often do succeed, and over the lastfew years, I have studied when and why this happens. In this talk, I willsurvey some recent work in this area covering optimization theory, algorithms,and applications in signal processing and machine learning. In the process, wewill encounter a surprisingly rich array of mathematical tools spanningnonsmooth analysis, semi-algebraic geometry, and high dimensional probabilityand statistics.</p><p><span>Biography: Dmitriy Drusvyatskiy received his PhD from theOperations Research and Information Engineering department at CornellUniversity in 2013, followed by a post doctoral appointment in theCombinatorics and Optimization department at Waterloo, 2013-2014. He joined theMathematics department at University of Washington as an Assistant Professor in2014, and was promoted to an Associate Professor in 2019, and to Full Professorin 2022. Dmitriy's research broadly focuses on designing and analyzing algorithmsfor large-scale optimization problems, primarily motivated by applications indata science. Dmitriy has received a number of awards, including the Air ForceOffice of Scientific Research (AFOSR) Young Investigator Program (YIP) Award,NSF CAREER, SIAG/OPT Best Paper Prize 2023, Paul Tseng Faculty fellowship2022-2026, INFORMS Optimization Society Young Researcher Prize 2019, andfinalist citations for the Tucker Prize 2015 and the Young Researcher BestPaper Prize at ICCOPT 2019. Dmitriy is currently a co-PI of the NSF fundedTransdisciplinary Research in Principles of Data Science (TRIPODS) institute atUniversity of Washington.</span></p><p></p><p> </p>

Tuesday, February 27, 2024 - 17:00 to 18:00

Clark, 110