MARL Chapter 9.8.3: Alpha Zero Introduction

This meeting will continue the material from Chapter 9 in Multi-Agent Reinforcement Learning: Foundations and Modern Approaches. Last time we covered Monte Carlo Tree search and showed how it can be extended to self-play turn taking games. This meeting we will lay the foundation needed to incorporate deep learning into MCTS and how that can be applied to perform policy improvement. We will discuss AlphaZero but also some recent refinements that change the way the policy function is used to direct tree search with Gumbel sampling.

As usual you can find below links to the textbook, previous chapter notes, slides, and recordings of some of the previous meetings.

Meetup Links:
Recordings of Previous RL Meetings
Recordings of Previous MARL Meetings
Short RL Tutorials
My exercise solutions and chapter notes for Sutton-Barto
My MARL repository
Kickoff Slides which contain other links
MARL Kickoff Slides

MARL Links:
Multi-Agent Reinforcement Learning: Foundations and Modern Approaches
MARL Summer Course Videos
MARL Slides

Sutton and Barto Links:
Reinforcement Learning: An Introduction by Richard S. Sutton and Andrew G. Barto
Video lectures from a similar course