A Lower Bound for the Sample Complexity of Inverse Reinforcement Learning | Read Paper on Bytez