This is a chess game, with a full interactive GUI. You can play chess against an AI, or watch two AIs play each other in simulation.

This program plays chess by modeling chess as a Markov decision process. Each piece has a very specefic set of moves it can make given the state of the board. I expand this tree of all possible moves and then used a heuristic scoring function to approximate how likely/desirable each state is. This allows me to compute the state with the highest expected value four to six moves forward in the game tree. Once I have the most desirable state at a leaf in the tree, I simply need to backrack to the immediate move that leads to it.

Clone my repo and run javac userInterface.java then java UserInterface