Robust Reinforcement Learning using Least Squares Policy Iteration with Provable Performance Guarantees
This addresses the problem of learning policies robust to model-reality mismatches in reinforcement learning, representing an incremental improvement with theoretical guarantees.
This paper tackles robust reinforcement learning for Markov Decision Processes with parameter uncertainties by proposing Robust Least Squares Policy Iteration (RLSPI), which provides provable performance guarantees with a general weighted Euclidean norm bound on policy error and demonstrates performance on standard benchmarks.
This paper addresses the problem of model-free reinforcement learning for Robust Markov Decision Process (RMDP) with large state spaces. The goal of the RMDP framework is to find a policy that is robust against the parameter uncertainties due to the mismatch between the simulator model and real-world settings. We first propose the Robust Least Squares Policy Evaluation algorithm, which is a multi-step online model-free learning algorithm for policy evaluation. We prove the convergence of this algorithm using stochastic approximation techniques. We then propose Robust Least Squares Policy Iteration (RLSPI) algorithm for learning the optimal robust policy. We also give a general weighted Euclidean norm bound on the error (closeness to optimality) of the resulting policy. Finally, we demonstrate the performance of our RLSPI algorithm on some standard benchmark problems.