報告題目: Some Recent Progress in Deep Reinforcement Learning

報告地點: 437bwin必贏國際官網B404會議室

報告時間: 2025年2月24日上午9點

報告人: Paul Weng

報告人單位: Duke Kunshan University

報告人簡介：

Paul Weng, a tenured associate professor at Duke Kunshan University, was previously an associate professor at the UM-SJTU Joint Institute and held regular or visiting faculty positions at multiple universities. As a top-tier AI researcher, he regularly publishes in top venues (e.g., IJCAI, AAAI, ICML, ICLR…) and has served as an area chair at AAAI and ECAI. He has received best paper awards of MIWAI and ALA.

His main research work lies in AI and machine learning, with focuses on adaptive control (reinforcement learning, Markov decision process), multi-objective optimization (compromise programming, fair optimization), and preference handling (representation, elicitation, and learning).

報告摘要：

Deep reinforcement learning (RL) is a generic and powerful machine learning approach to solve sequential decision-making or control problems. This talk will present an overview of the different research directions explored in our team (i.e., fair RL, exploitation of equivariance, RL from human feedback, application to routing problems) and the results we achieved in these directions.

邀請人：王皓