Batch policy learning in average reward Markov decision processes
Peng Liao, Zhengling Qi, Runzhe Wan, Predrag Klasnja, Susan A. Murphy
The Annals of Statistics · 2022 · Volume 50 · Issue 6
Abstract
We consider the batch (off-line) policy learning problem in the infinite horizon Markov decision process. Motivated by mobile health applications, we focus on learning a policy that maximizes the long-term average reward. We propose a doubly robust estimator for the average reward and show that it achieves semiparametric efficiency. Further, we develop an optimization algorithm to compute the optimal policy in a parameterized stochastic policy class. The performance of the estimated policy is measured by the difference between the optimal average reward in the policy class and the average reward of the estimated policy and we establish a finite-sample regret guarantee. The performance of the method is illustrated by simulation studies and an analysis of a mobile health study promoting physical activity.
Keywords
62G05, Average reward, doubly robust estimator, Markov decision process, policy optimization
Enable JavaScript for citation tools and additional paper interactions.