Research/Blog
CellStrat > Research/Blog > Industry > Gaming > A Summary of AlphaZero – Training an agent to play classical games Tabula Rasa
A Summary of AlphaZero – Training an agent to play classical games Tabula Rasa
- July 21, 2021
- Posted by: vsinghal
- Category: Gaming papers Reinforcement Learning
No Comments
Author : Shubha Manikarnike
Submitted on : 5 Sep 2020
Abstract – This paper explains the AlphaGoZero algorithm which was developed by Google Deepmind. It achieved super human performance in the Game of GO. It discusses Monte Carlo Tree Search (MCTS) algorithm which is integrated with an Actor Critic approach. The agent learns how to play the game Tabula Rasa by playing against itself.
Keywords – AlphaGo, AlphaZero, Monte Carlo Tree Search, Tabula Rasa