Imitation Learning in Discounted Linear MDPs without exploration assumptions

Open in new window