Courses/Reinforcement Learning

Reinforcement Learning

Policy gradient, Q-learning, and actor-critic methods built from the Bellman equation up. Train agents in MuJoCo and Atari.

Intermediate14 weeks · 18 lessons

Your progress00 / 18 · 0%

4 lessons here are on the Guided Foundations Path· Steps 11-14

Lessons

W01Reinforcement Learning: Agents, Rewards, and MDPsGuided Path · Beginner W02Bandits: Exploration, Exploitation, and RegretGuided Path · Beginner W03Returns, Value, and the Bellman EquationGuided Path · Beginner W04Bandits: UCB1 and Thompson SamplingGuided Path · Beginner W01Week 1: Reinforcement Learning Problem FormulationGeneral W02Week 2: Multi-Armed BanditsGeneral W03Week 3: Dynamic Programming for Finite MDPsGeneral W04Week 4: Monte Carlo and Temporal-Difference LearningGeneral W05Week 5: Function Approximation in Reinforcement LearningGeneral W06Week 6: Deep Q-Learning and VariantsGeneral W07Week 7: Policy Gradient and Actor–Critic MethodsGeneral W08Week 8: Modern Deep Reinforcement Learning AlgorithmsGeneral W09Week 9: Exploration, Partial Observability, and Multi-Agent Reinforcement LearningGeneral W10Week 10: Model-Based Reinforcement Learning and PlanningGeneral W11Week 11: Offline Reinforcement LearningGeneral W12Week 12: Reinforcement Learning from Human FeedbackGeneral W13Week 13: Direct Preference Optimization and GRPOGeneral W14Week 14: Agentic Systems and Course CapstoneGeneral