Learning from humans: what is inverse reinforcement learning?