Game-Theoretic Robust Reinforcement Learning Handles Temporally-Coupled Perturbations
This addresses robustness in RL for real-world deployments where uncertainties are time-coupled, offering a novel solution to a specific bottleneck.
The paper tackles the problem of reinforcement learning systems needing robustness to temporally-coupled perturbations, which are uncertainties linked across time, and proposes GRAD, a game-theoretic approach that achieves higher robustness in experiments on continuous control tasks compared to prior methods.
Deploying reinforcement learning (RL) systems requires robustness to uncertainty and model misspecification, yet prior robust RL methods typically only study noise introduced independently across time. However, practical sources of uncertainty are usually coupled across time. We formally introduce temporally-coupled perturbations, presenting a novel challenge for existing robust RL methods. To tackle this challenge, we propose GRAD, a novel game-theoretic approach that treats the temporally-coupled robust RL problem as a partially observable two-player zero-sum game. By finding an approximate equilibrium within this game, GRAD optimizes for general robustness against temporally-coupled perturbations. Experiments on continuous control tasks demonstrate that, compared with prior methods, our approach achieves a higher degree of robustness to various types of attacks on different attack domains, both in settings with temporally-coupled perturbations and decoupled perturbations.