Is Reinforcement Learning More Difficult Than Bandits? A Near-optimal Algorithm Escaping the Curse of Horizon.

Zihan Zhang, Xiangyang Ji, Simon S. Du

Browse the full COLT paper archive.