【学术报告】Online Decision Making: Algorithm, Regret, Constraints and Uncertainty（夏冬-香港科技大学）

【主讲人简介】：Dr. Dong XIA is an Associate Professor in Department of Mathematics at Hong Kong University of Science and Technology. He was a Postdoctoral Research Scientist in Department of Statistics at Columbia University and a Visiting Assistant Professor in Department of Statistics at University of Wisconsin-Madison. He obtained his Ph.D. in Computational Science and Engineering and Mathematics from Georgia Institute of Technology in 2016. His research interest lies in high-dimensional statistics, machine learning theory and optimization.

【内容简介】：We will discuss online decision-making problems in scenarios where covariates are high-dimensional or personalized covariates are unavailable. Our focus is on the ε‌-greedy algorithm for decision making and online gradient descent for estimating model parameters. By carefully balancing exploration and exploitation, we achieve a trade-off between regret performance and estimation accuracy. Additionally, we explore online decision-making under constraints (such as knapsack problems) within a primal-dual framework, demonstrating that sublinear regret is achievable. Finally, we propose an online debiasing approach based on inverse propensity weighting (IPW) for uncertainty quantification. Real data examples will also be discussed.

【讲座时间】：2026年6月5日（星期五）上午10点

【讲座地点】：文科综合楼1801会议室