On the Global Convergence of Policy Gradient in Average Reward Markov Decision Processes

Re-assign community