56. Reinforcement learning

Covers agents, environments, rewards, policies, value functions, exploration, and offline evaluation risks. Learners see when reinforcement learning fits decision problems and when simpler methods are safer.