This document provides an overview of various reinforcement learning methods including policy-based methods like REINFORCE and TRPO, value-based methods like Q-Learning and Sarsa, model-based methods involving planning, imitation learning methods like DAGGER and GAIL, and evolutionary algorithms like CEM. It also briefly discusses considerations for these methods like advantages of being more data efficient than policy-based RL but having difficulties with high-dimensional or continuous action spaces and stochastic policies.