On the Value of Interaction and Function Approximation in Imitation Learning

Open in new window