Imitation Learning trains an agent to mimic expert demonstrations rather than learning solely from rewards. It’s useful when designing a reward function is difficult or dangerous. Behavioral cloning and inverse reinforcement learning are common approaches in imitation learning.