Distributionally Robust Off-Dynamics Reinforcement Learning: Provable Efficiency with Linear Function Approximation

TL;DR

This paper introduces a new algorithm for off-dynamics reinforcement learning that is robust to model uncertainties, providing provable efficiency guarantees with linear function approximation and demonstrating strong empirical performance.

Contribution

It develops the first provably efficient online distributionally robust RL algorithm with linear function approximation for off-dynamics scenarios, addressing nonlinearity issues in the dual formulation.

Findings

The proposed DR-LSVI-UCB algorithm achieves polynomial suboptimality bounds.

Using a $d$-rectangular uncertainty set removes nonlinearity and error propagation.

Numerical experiments confirm the robustness and effectiveness of the method.

Abstract

We study off-dynamics Reinforcement Learning (RL), where the policy is trained on a source domain and deployed to a distinct target domain. We aim to solve this problem via online distributionally robust Markov decision processes (DRMDPs), where the learning algorithm actively interacts with the source domain while seeking the optimal performance under the worst possible dynamics that is within an uncertainty set of the source domain's transition kernel. We provide the first study on online DRMDPs with function approximation for off-dynamics RL. We find that DRMDPs' dual formulation can induce nonlinearity, even when the nominal transition kernel is linear, leading to error propagation. By designing a $d$-rectangular uncertainty set using the total variation distance, we remove this additional nonlinearity and bypass the error propagation. We then introduce DR-LSVI-UCB, the first provably efficient online DRMDP algorithm for off-dynamics RL with function approximation, and establish a polynomial suboptimality bound that is independent of the state and action space sizes. Our work makes the first step towards a deeper understanding of the provable efficiency of online DRMDPs with linear function approximation. Finally, we substantiate the performance and robustness of DR-LSVI-UCB through different numerical experiments.