Sample Efficient Imitation Learning via Reward Function Trained in Advance

Lihua Zhang

Imitation learning (IL) is a framework that learns to imitate expert behavior from demonstrations. Recently, IL shows promising results on high dimensional and control tasks. Howev…