Accelerating Primal-dual Methods for Regularized Markov Decision Processes - 42Papers