Publication

Imitation Learning in Discounted Linear MDPs without exploration assumptions