What is Reinforcement Learning?Taxonomy of RL Methods History of Reinforcement Learning

Multi-Armed Bandits MDP and Dynamic Programming Monte Carlo and Temporal Difference Q-Learning DQN DQN Improvements

Policy Gradient Theorem and REINFORCE TRPO PPO RL for Sequence Generation and RLHF

Actor-Critic Framework DDPG TD3 and SAC

Dyna and Learned Models Model Predictive Control AlphaZero and MuZero

Imitation Learning Offline RL Exploration Goal-Conditioned RL Multi-Agent RL

Value-Based Methods

DQN Improvements

Double DQN, Dueling DQN, Prioritized Experience Replay, and Rainbow.

Placeholder content for DQN Improvements.

DQN

Deep Q-Networks with experience replay and target networks.

Policy Gradient Theorem and REINFORCE

Deriving the policy gradient and the REINFORCE Monte Carlo estimator.