A Reinforcement Learning Approach for Solving Chess Endgames

In this project we show how a Reinforcement Learning approach can be successfully applied in chess. This is done by focusing on KRK endgames and by implementing the Q-Learning algorithm with different exploration policies. The main goal of this research was to train an artificial agent able to win the endgames as a White Player against an "experienced" Black Player.

In order to train an agent be sure to follow the following steps:

1. Installed python3 and all required packages, in particular "cocos" if you want the fancy GUI
2. Create a /res dir in src dir.
3. Launch BaseParams.py in order to build the empty Q-Matrix
4. To train the Agent using the Q-Learning algorithm set necessary parameters at the end of QLearning.py, and of course launch it.
5. To see statistcs relative to how many wins and the average number of moves regarding the White Player launch StatsQ.py with relative parameters
6. If you want to see a command line GUI change Play(fp,TRUE)
7. If you want to see the fancy GUI launch GUI from res /GUI being sure to use the correct memory file in the main

Have fun from Matthia, Yaroslav, Zacharias and Evangelos!!!!!

@paintception @YAR0N @zacharias33 @Angelo404