Timezone: »

Offline Reinforcement Learning with Closed-Form Policy Improvement Operators
Jiachen Li · Edwin Zhang · Ming Yin · Jerry Bai · Yu-Xiang Wang · William Wang

Thu Jul 27 01:30 PM -- 03:00 PM (PDT) @ Exhibit Hall 1 #112
Event URL: https://cfpi-icml23.github.io/ »

Behavior constrained policy optimization has been demonstrated to be a successful paradigm for tackling Offline Reinforcement Learning. By exploiting historical transitions, a policy is trained to maximize a learned value function while constrained by the behavior policy to avoid a significant distributional shift. In this paper, we propose our closed-form policy improvement operators. We make a novel observation that the behavior constraint naturally motivates the use of first-order Taylor approximation, leading to a linear approximation of the policy objective. Additionally, as practical datasets are usually collected by heterogeneous policies, we model the behavior policies as a Gaussian Mixture and overcome the induced optimization difficulties by leveraging the LogSumExp's lower bound and Jensen's Inequality, giving rise to a closed-form policy improvement operator. We instantiate both one-step and iterative offline RL algorithms with our novel policy improvement operators and empirically demonstrate their effectiveness over state-of-the-art algorithms on the standard D4RL benchmark. Our code is available at https://cfpi-icml23.github.io/.

Author Information

Jiachen Li (University of California, Santa Barbara)
Edwin Zhang (Harvard)
Ming Yin (UCSB/Princeton)
Jerry Bai (Horizon Robotics)
Yu-Xiang Wang (UC Santa Barbara / Amazon)
William Wang (UCSB)

More from the Same Authors