Monte Carlo Tree Search and Machine Learning Techniques on Block Go Programs

Black Go, a game similar to Go, uses Tetris-like pieces and can teach complex Go logic through a fun and easy game process. Monte Carlo tree search can be used to design Block Go programs, and machine learning techniques can train game records and obtain useful game patterns. The paper discusses how to develop a pattern database on Block Go and its performance on Monte Carlo tree search to create a good Block Go program.