Robust Q-Learning under Corrupted Rewards

Sreejeet Maity, Aritra Mitra

Recently, there has been a surge of interest in analyzing the non-asymptotic behavior of model-free reinforcement learning algorithms. However, the performance of such algorithms i…