AQFC2015

Reinforcement Learning in Dynamic Games with Imperfect Information: Challenges and Opportunities

----------------------------------------------------------------------------------------------------
     Department of Systems Engineering and Engineering Management
                             The Chinese University of Hong Kong
----------------------------------------------------------------------------------------------------
Date: Friday, March 13, 2026, 10:30am to 11:30am HKT
Venue: ERB 513, The Chinese University of Hong Kong
Title: Reinforcement Learning in Dynamic Games with Imperfect Information: Challenges and Opportunities
Speaker: Dr. LI Tao, City University of Hong Kong
 
 
Abstract:
 
Imperfect information games model scenarios in which players do not know the exact state of the game or the history of moves, offering a principled mathematical framework for decision-making over complex networks where networked agents have only partial observations of the system state. Despite their broader relevance, imperfect information games have had limited applications due to the challenges in solving for equilibrium strategies under agents' sequential rationality. This talk discusses new directions in data-driven equilibrium-seeking, particularly in unknown game environments, enabled by recent advances in reinforcement learning (RL). We highlight our contributions to both model-based and model-free RL in imperfect-information stochastic games grounded in optimal control theory. Finally, we conclude this talk by discussing the promising future directions and opportunities.
 
Biography:
 
Dr. Tao LI received a B.S. in Mathematics from Xiamen University, Fujian, China, in 2018 and a Ph.D. in Electrical and Computer Engineering from New York University in 2025. He was a visiting scholar at IBM Research Thomas J. Watson Research Center before joining City University of Hong Kong, where he is currently an assistant professor in the Department of Systems Engineering. His research interests include game theory, control and optimization, and reinforcement learning with applications to secure and resilient cyber-physical network system design, defense, and management.
 
Everyone is welcome to attend the talk!
SEEM-5202 Website: http://seminar.se.cuhk.edu.hk
Date: 
Wednesday, March 13, 2026 - 10:30 to 11:30
 
Date: 
Friday, March 13, 2026 - 10:30