Model-Free Robust $φ$-Divergence Reinforcement Learning Using Both Offline and Online Data
This work addresses robustness in reinforcement learning for applications where simulators differ from real-world settings, offering incremental improvements in handling data distribution assumptions.
The paper tackles robust reinforcement learning under model uncertainty by proposing model-free algorithms for learning optimal robust policies using offline and hybrid data, achieving theoretical guarantees for high-dimensional systems with general function approximation.
The robust $φ$-regularized Markov Decision Process (RRMDP) framework focuses on designing control policies that are robust against parameter uncertainties due to mismatches between the simulator (nominal) model and real-world settings. This work makes two important contributions. First, we propose a model-free algorithm called Robust $φ$-regularized fitted Q-iteration (RPQ) for learning an $ε$-optimal robust policy that uses only the historical data collected by rolling out a behavior policy (with robust exploratory requirement) on the nominal model. To the best of our knowledge, we provide the first unified analysis for a class of $φ$-divergences achieving robust optimal policies in high-dimensional systems with general function approximation. Second, we introduce the hybrid robust $φ$-regularized reinforcement learning framework to learn an optimal robust policy using both historical data and online sampling. Towards this framework, we propose a model-free algorithm called Hybrid robust Total-variation-regularized Q-iteration (HyTQ: pronounced height-Q). To the best of our knowledge, we provide the first improved out-of-data-distribution assumption in large-scale problems with general function approximation under the hybrid robust $φ$-regularized reinforcement learning framework. Finally, we provide theoretical guarantees on the performance of the learned policies of our algorithms on systems with arbitrary large state space.