mlisthenewcool/_M2_reinforcement-learning

Reinforcement Learning (RL)

This repository provides code, links and reflexions of various methods to learn from interaction in order to achieve goals. Ideas and structure relies heavily on theoretical materials from Richard S. Sutton and Andrew G. Barto's Reinforcement Learning book (2nd edition).

All code is written in Python 3.6+. Some environments are provided by OpenAI Gym.

Table of contents

1. Tabular solution methods

• Multi-armed bandits
• Multi-armed contextual bandits
• Monte Carlo Tree Search
• Temporal-Difference Learning