Competitive Experience Replay

Hao Liu, Alexander Trott, Richard Socher +1

Deep learning has achieved remarkable successes in solving challenging reinforcement learning (RL) problems when dense reward function is provided. However, in sparse reward enviro…