Computational Game Theory & Machine Learning: Applications To The Game Of Sim
Harry Lewis Bryant · UA Campus Repository (The University of Arizona) · 2019
This paper focuses on the process of creating bots that use the Min-Max algorithm and Alpha-Beta pruning in conjunction with different evaluation functions to play the game of Sim. This process strengthens our understanding of computational game theory by involving the usage of these algorithms and our evaluation functions. Also discussed in this paper is a review of the basics of game theory, including information on winning and losing positions, as well as an explanation of the evaluation functions used. In addition, the rules of the game of Sim are explained, given that Sim is the particular game that this project of machine learning is based on. The latest stage of the project is comprised of a bot that uses a heuristic evaluation function constructed using a multi-layered neural network.