Provably Efficient Offline Reinforcement Learning with Perturbed Data Sources.

Chengshuai Shi, Wei Xiong, Cong Shen, Jing Yang

Browse the full ICML paper archive.