/EEA-Opponent-Modeling

Using Konane to demonstrate the possible strength of the Estimation Exploration Algorithm for opponent modeling

Primary LanguageOpenEdge ABLGNU General Public License v2.0GPL-2.0

EEA-Opponent-Modeling

This is the repository for my senior year thesis.

The basic research question I am exploring is, "How good is the Estimation Exploration Algorithm for opponent modeling?"

I will be exploring this question by using the EEA to try to accurately model an opponent's static evaluator in a board game. That means that the EEA will be using a game's moves as a means of probing the hidden system that is the opponent. The hope is that, over time the EEA will be able to evolve the static evaluator of a game engine to fit to what the opponent is doing so that it may predict the opponent's next move accurately.

I'm currently working with the game Konane and making very good progress! If you want to see what I'm working on at the moment, go to the Konane directory. This has all of the code and data that I've been working on since I started the project.