The seminar meets Friday, 2:00-5:00 pm (UTC+8) in a online-offline mixed mode. The offline venue is located in 501 Daoyuan Building. We also provide online zoom meeting room for the convenience of discussion and replay. The zoom link will be sent by email. If you are interested, please contact us.

Date	Speaker and Title
Jun 11	Ziniu Li (CUHK-SZ) Global optimality guarantees for policy gradient methods
Jun 18	Xuhui Liu (NJU) Universal Off-Policy Evaluation
Jun 25	Tian Xu (NJU) Fundamental limits of imitation learning with known transition
Jul 02	Liangqi Liu (CUHK-SZ) Reinforcement learning via sequence modeling - Beyond Markovian assumption
Jul 09	Hao Liang (CUHK-SZ) Revisit minimax lower bounds in episodic rl in finite mdps
Jul 16	Yingru Li (CUHK-SZ) On the worst-case regret of Linear Thompson Sampling
Jul 23	Ziniu Li (CUHK-SZ) Optimality of Policy Gradient Methods
Jul 30	Tian Xu (NJU) Provably Efficient Exploration in Policy Optimization
Aug 06	Guiyu Hong (CUHK-SZ) Simple Bayesian Algorithms for Best-Arm Identification
Aug 13	Xiuwen Wang (CUHK-SZ) Provably Efficient Reinforcement Learning with Aggregated States