Ziniu Li
About me
I am a Ph.D. student at The Chinese University of Hong Kong, Shenzhen (CUHKSZ), advised by Prof. Zhi-Quan (Tom) Luo.
My research interests include reinforcement learning theory, algorithms, and applications.
I have worked/interned at Tencent, Nanjing University, Cardinal Operations, etc.
My curriculum vitae can be downloaded from here.
Feel free to contact me if you want to discuss some ideas.
Selected Work
*: indicating equal contribution or alphabetic ordering.
Theoretical Analysis of Offline Imitation With Supplementary Dataset
Ziniu Li* , Tian Xu*, Yang Yu, Zhi-Quan Luo
arXiv:2301.11687
(This work explores the framework of transfer learning (or domain adaption) used in CV, NLP, and RL, in terms of data coverage, importance sampling (or re-weighting), and feature alignment, answering questions including what kind of source data is helpful? when can the importance sampling improve a lot?)
Rethinking ValueDice: Does It Really Improve Performance?
Ziniu Li* , Tian Xu*, Yang Yu, Zhi-Quan Luo
In Proceedings of the 10th International Conference on Learning Representations (ICLR) (Blog Track), 2022
(This work presents the first reduction of offline AIL to BC through a new DP-based analysis, proving that AIL is no better than BC in the offline setting and verifying this by experiments)
Service
Reviewer
NeurIPS’2022 (Top Reviewer), ICML’2022 (Outstanding Reviewer), ICLR’2022 (Highlighted Reviewer), DAI’2020.
Teaching Assistant
Lecturer
Award
|