Using Tic-Tac-Toe for Learning Data Mining Classifications and Evaluations

Abstract

Tic-Tac-Toe game is a popular two-player game played on a three by three grid. The first objective of this study is to examine whether machine learners can successfully classify Tic-Tac-Toe finished games. The second objective is to find out whether novices learning data mining classifications can successfully conduct experiments to evaluate the performance of machine learners using different evaluation methods. Seven machine learners are used and three evaluation methods are applied in this study. The experimental results show that machine classifiers can successfully judge the finished games and novice can correctly conduct evaluations. Also, in-depth instructional pedagogy further improves the correctness of evaluations.