Averaged-DQN: Variance Reduction and Stabilization for Deep Reinforcement Learning.

Oron Anschel, Nir Baram, Nahum Shimkin

Browse the full ICML paper archive.