A Comparism of the Performance of Supervised and Unsupervised Machine Learning Techniques in evolving Awale/Mancala/Ayo Game Player

TL;DR

This paper compares supervised and unsupervised machine learning techniques in developing Awale game players, evaluating their performance with various strategies to identify the most effective approach.

Contribution

It provides a comparative analysis of supervised and unsupervised learning methods for evolving Awale game agents, highlighting their relative strengths and weaknesses.

Findings

Supervised learning techniques outperform unsupervised methods in certain conditions.

Combining minimax with endgame databases enhances game-playing performance.

The study identifies the most effective machine learning approach for Awale game agents.

Abstract

Awale games have become widely recognized across the world, for their innovative strategies and techniques which were used in evolving the agents (player) and have produced interesting results under various conditions. This paper will compare the results of the two major machine learning techniques by reviewing their performance when using minimax, endgame database, a combination of both techniques or other techniques, and will determine which are the best techniques.