Foundations & Dynamic Programming
A 14-module path through foundations & dynamic programming, in study order from foundations to the frontier.
Modules
- The RL Problem and the Agent-Environment Loop
- Rewards and Returns
- The Markov Property
- Markov Decision Processes
- Policies and Value Functions
- State-Value and Action-Value Functions
- The Bellman Equations
- The Bellman Optimality Equation
- Dynamic Programming for MDPs
- Iterative Policy Evaluation
- Policy Iteration
- Value Iteration
- The Exploration-Exploitation Dilemma
- Multi-Armed Bandits and Regret