Deep Reinforcement Learning and Rainbow

Recently, I’ve been interested in learning more about reinforcement learning. The impetus for this presentation is this paper, Revisiting Rainbow, that came out at ICML this year, 2021. In it, the authors build on some earlier breakthroughs, so I decided that this would be a good time to review those earlier breakthroughs in more depth, so I could understand this paper better.

Revisiting Rainbow, Part 1 – Introduction

Revisiting Rainbow, Part 2 – Q Learning Basics

Revisiting Rainbow, Part 3 – Deep Q Network

Revisiting Rainbow, Part 4 – Prioritized Experience Replay

Revisiting Rainbow, Part 5 – Multi-Step Learning

Revisiting Rainbow, Part 6 – Double Q Learning

Revisiting Rainbow, Part 7 – Dueling Networks

Revisiting Rainbow, Part 8 – Noisy Nets

Revisiting Rainbow, Part 9 – Distributional Reinforcement Learning

Revisiting Rainbow, Part 10 – Rainbow

Revisiting Rainbow, Part 11 – Revisiting Rainbow