報告題目：Algorithmic Design for Wassernstein Distributionally Robust Optimization in Machine Learning
報告時間：2020年10月28日（周三）10:15-11:15
報告地點：北辰校區理學院（西教五）416
報告嘉賓：陳彩華 副教授（南京大學）
報告摘要：Wasserstein Distributionally Robust Stochastic Optimization (DRSO) is concerned with finding decisions that perform well on data that are drawn from the worstcase probability distribution within a Wasserstein ball centered at a certain nominal distribution. In recent years, it has been shown that various DRSO formulations of learning models admit tractable convex reformulations. However, most existing works propose to solve these convex reformulations by general-purpose solvers, which are not well suited for tackling largescale problems.
In this talk,we focus on Wasserstein distributionally robust support vector machine (DRSVM) problems and logistic regression (DRLR) problems, and propose two novel first order algorithms to solve them. The updates in each iteration of these algorithms can be computed in a highly efficient manner. Our numerical results indicate that the proposed methods are orders of magnitude faster than the state-of-the-art, and the performance gap grows considerably as the problem size increases.Advanced models such as robust classifi-cation with fairness and unlabelled data are also discussed.
嘉賓簡介：陳彩華，副教授，南京大學理學博士，新加坡國立大學聯合培養博士，曾赴新加坡國立大學、香港中文大學、香港理工大學、香港浸會大學等學習與訪問。主持/完成的基金包括國家自然科學基金面上項目、青年項目，江蘇省自然科學基金面上項目、青年項目，參與國家自然科學基金重點項目，代表作發表在Mathematical Programming，SIAM Journal on Optimization,SIAM Journal on Imaging Science及CVPR、NIPS等國際知名學術期刊與會議，其中多篇論文入選ESI高被引論文。獲華人數學家聯盟最佳論文獎（2017、2018連續兩年），中國運籌學會青年科技獎（2018），南京大學青年五四獎章(2019)，入選首批南京大學仲英青年學者（全校10人，2018）及江蘇省社科優青（2019）。
