Tic-tac-toe prediction based on machine learning methods

Abstract

With the development of artificial intelligence (AI) in modern society, more and more research findings related to AI are being discovered. Machine learning allows for automation and mechanical control compared to traditional manual control. This article attempts to implement simulation and analysis of specific problems using some specific modules in python code. Throughout the research process, three different algorithmic models, namely Random Forest, Decision Tree, and SVM, were chosen and used to play Tic Tac Toe. By using mathematical combinations of sequences that can be played by the system, different artificial neural networks are trained, and the training results are compared and analyzed to ultimately find the most suitable algorithmic model. The choice of algorithmic model can contribute to a more accurate prediction of tic-tac-toe subsequently. Through the training and prediction of data sets, it is considered that SVM in some cases have the strongest performance among others. The model may be more applicable to the prediction and analysis of tic-tac-toe games. By analogy, the model should be able to help find more excellent performances in tic-tac-toe games.