Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm

This paper focuses on the development and performance of the AlphaZero algorithm and how it achieves tabula rasa learning, that is learning from a blank slate or without any domain specific knowledge.

AlphaZero uses the Monte-Carlo Tree search algorithm to train the neural network through self play reinforcement learning. AlphaZero, without any game specific knowledge is able to learn the games of Chess, Shogi and Go. Its able to match and even outperform other models, Stockfish (chess), Elmo (shogi) and AlphaGo Zero(Go) which were trained specifically for these games. You can find the original paper here