Deep Reinforcement Learning for 5*5 Multiplayer Go
This work addresses the challenge of extending AI techniques to multiplayer games, but it is incremental as it adapts existing methods like AlphaZero to a new variant.
The paper tackled the problem of applying deep reinforcement learning and search algorithms to multiplayer Go, showing that these methods improved the level of play in a 5x5 game with more than two players, though no concrete numbers were provided.
In recent years, much progress has been made in computer Go and most of the results have been obtained thanks to search algorithms (Monte Carlo Tree Search) and Deep Reinforcement Learning (DRL). In this paper, we propose to use and analyze the latest algorithms that use search and DRL (AlphaZero and Descent algorithms) to automatically learn to play an extended version of the game of Go with more than two players. We show that using search and DRL we were able to improve the level of play, even though there are more than two players.