Reinforcement learning by reward-weighted regression for operational space control.

Jan Peters, Stefan Schaal

Browse the full ICML paper archive.