The Utility of Sparse Representations for Control in Reinforcement Learning

Open in new window