Optimality and Approximation with Policy Gradient Methods in Markov Decision Processes.

Alekh Agarwal, Sham M. Kakade, Jason D. Lee, Gaurav Mahajan

Browse the full COLT paper archive.