Saved in:
Bibliographic Details
Main Authors: Wang, Shengbo, Si, Nian, Blanchet, Jose, Zhou, Zhengyuan
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2302.13203
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909275113127936
author Wang, Shengbo
Si, Nian
Blanchet, Jose
Zhou, Zhengyuan
author_facet Wang, Shengbo
Si, Nian
Blanchet, Jose
Zhou, Zhengyuan
contents We consider a reinforcement learning setting in which the deployment environment is different from the training environment. Applying a robust Markov decision processes formulation, we extend the distributionally robust $Q$-learning framework studied in Liu et al. [2022]. Further, we improve the design and analysis of their multi-level Monte Carlo estimator. Assuming access to a simulator, we prove that the worst-case expected sample complexity of our algorithm to learn the optimal robust $Q$-function within an $ε$ error in the sup norm is upper bounded by $\tilde O(|S||A|(1-γ)^{-5}ε^{-2}p_{\wedge}^{-6}δ^{-4})$, where $γ$ is the discount rate, $p_{\wedge}$ is the non-zero minimal support probability of the transition kernels and $δ$ is the uncertainty size. This is the first sample complexity result for the model-free robust RL problem. Simulation studies further validate our theoretical results.
format Preprint
id arxiv_https___arxiv_org_abs_2302_13203
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle A Finite Sample Complexity Bound for Distributionally Robust Q-learning
Wang, Shengbo
Si, Nian
Blanchet, Jose
Zhou, Zhengyuan
Machine Learning
We consider a reinforcement learning setting in which the deployment environment is different from the training environment. Applying a robust Markov decision processes formulation, we extend the distributionally robust $Q$-learning framework studied in Liu et al. [2022]. Further, we improve the design and analysis of their multi-level Monte Carlo estimator. Assuming access to a simulator, we prove that the worst-case expected sample complexity of our algorithm to learn the optimal robust $Q$-function within an $ε$ error in the sup norm is upper bounded by $\tilde O(|S||A|(1-γ)^{-5}ε^{-2}p_{\wedge}^{-6}δ^{-4})$, where $γ$ is the discount rate, $p_{\wedge}$ is the non-zero minimal support probability of the transition kernels and $δ$ is the uncertainty size. This is the first sample complexity result for the model-free robust RL problem. Simulation studies further validate our theoretical results.
title A Finite Sample Complexity Bound for Distributionally Robust Q-learning
topic Machine Learning
url https://arxiv.org/abs/2302.13203