Near-Optimal Sample Complexities of Divergence-based S-rectangular Distributionally Robust Reinforcement Learning

TL;DR

This paper establishes near-optimal sample complexity bounds for divergence-based S-rectangular distributionally robust reinforcement learning, highlighting improved theoretical understanding and practical performance in robust RL models.

Contribution

It provides the first near-optimal sample complexity results for divergence-based S-rectangular DR-RL, matching optimal dependence on key parameters.

Findings

Sample complexity bounds of $ ilde{O}(|S||A|(1-gamma)^{-4}varepsilon^{-2})$ established.

Numerical experiments confirm fast learning performance of the proposed algorithm.

Results demonstrate the effectiveness of S-rectangular models in robust RL applications.

Abstract

Distributionally robust reinforcement learning (DR-RL) has recently gained significant attention as a principled approach that addresses discrepancies between training and testing environments. To balance robustness, conservatism, and computational traceability, the literature has introduced DR-RL models with SA-rectangular and S-rectangular adversaries. While most existing statistical analyses focus on SA-rectangular models, owing to their algorithmic simplicity and the optimality of deterministic policies, S-rectangular models more accurately capture distributional discrepancies in many real-world applications and often yield more effective robust randomized policies. In this paper, we study the empirical value iteration algorithm for divergence-based S-rectangular DR-RL and establish near-optimal sample complexity bounds of $\widetilde{O}(|\mathcal{S}||\mathcal{A}|(1-\gamma)^{-4}\varepsilon^{-2})$, where $\varepsilon$ is the target accuracy, $|\mathcal{S}|$ and $|\mathcal{A}|$ denote the cardinalities of the state and action spaces, and $\gamma$ is the discount factor. To the best of our knowledge, these are the first sample complexity results for divergence-based S-rectangular models that achieve optimal dependence on $|\mathcal{S}|$, $|\mathcal{A}|$, and $\varepsilon$ simultaneously. We further validate this theoretical dependence through numerical experiments on a robust inventory control problem and a theoretical worst-case example, demonstrating the fast learning performance of our proposed algorithm.