Efficient Q-Learning and Actor-Critic Methods for Robust Average Reward Reinforcement Learning.

Yang Xu 0003, Swetha Ganesh, Vaneet Aggarwal