Non-Adversarial Inverse Reinforcement Learning via Successor Feature Matching

Open in new window