Reinforcement Learning (RL) is a machine learning technique that learns the optimal strategy to solve a task by trial and error, guided by positive or negative reward, rather than by explicit supervision.