feyil / Simple-Chess-AI-with-Minimax

Implementations of minimax and alpha_beta_minimax algorithm for Chess

Simple Chess AI with Minimax Algorithm

Introduction

• This repo includes my solution of the given homework(1/2) in the scope of the Artifical Intelligence(CENG461) course which is given as a technical elective in 2019-2020 Fall semester by Computer Engineering Department at Izmir Institute of Technology.

• (*)README.md file uses some parts of the official Homework Doc to better express the purpose of the Homework.

Problem*

• In this homework, you should code a program that plays chess. The UI is already coded for you, you are asked to code the program which takes a scenario (or starting board) and calculates the best possible move. The program should play against the user, first AI makes the move then user and it goes like this.

Approach*

• The chess game is based on making decision based on the current board at each step. So you should create a tree based on legal moves at each step, and expand the leaves based on legal moves at each leaf as well. Since this problem is exponentially complex to the number of possible moves, you are asked to search for until depth 5 (in other words, AI only thinks 5 move ahead like ”AI-user-AI-user-AI”). There is an algorithm called minimax, which you can use in your program, for search in turn-based decision systems. Also you need a value function to evaluate each moves and what parameters your function depends on is up to you.

Example Scenarios*

• There are several notations to represent chess boards. In this homework, we use Forsyth-Edwards Notation (FEN). The links are given below to clarify what the characters of notation stands for: https://en.wikipedia.org/wiki/Forsyth%E2%80%93Edwards_Notation https://www.chessgames.com/fenhelp.html

• The 1 move ahead scenarios:

• Checkmating: 8/5K1k/8/8/8/8/8/6Q1 w - - 0 1 8/8/8/8/k7/8/1R3K2/1R6 w - - 0 1 1Q6/8/8/2R3K1/8/8/8/k7 w - - 0 1

• Preventing from checkmate: 6k1/r4Qpp/8/8/8/4K3/8/8 b - - 0 1 2k2R2/1b1p4/3B2r1/8/4q3/6p1/ppp5/1k1r4 b - - 0 1

• The 3 move ahead of checkmating scenarios: 3Q4/8/8/2R3K1/8/8/k7/8 w - - 0 1 8/8/8/8/8/7k/4KR2/6R1 w - - 0 1 3K2R1/n3r3/5P1P/8/n7/pp6/kp6/3Q4 b - - 0 1

Implementation and Result Showcase

• The implementation of chess_ai.py file provided by me as a solution. This file includes neccessary functions. I implemented two version of minimax algorithm with function names:

• minimax_ai_play(board, depth, maximizingPlayer)
• alpha_beta_minimax(board, depth, alpha, beta, maximizingPlayer)

• I have used below pseudocodes which can be found in (https://en.wikipedia.org/wiki/Minimax and https://en.wikipedia.org/wiki/Alpha%E2%80%93beta_pruning). The main purpose was to understand the core mechanism of the algorithms. Therefore implementations made far away from production readiness. In this aspect, I'm always open to feedback.

function minimax(node, depth, maximizingPlayer) is
    if depth = 0 or node is a terminal node then
        return the heuristic value of node
    if maximizingPlayer then
        value := −∞
        for each child of node do
            value := max(value, minimax(child, depth − 1, FALSE))
        return value
    else (* minimizing player *)
        value := +∞
        for each child of node do
            value := min(value, minimax(child, depth − 1, TRUE))
        return value

function alphabeta(node, depth, α, β, maximizingPlayer) is
    if depth = 0 or node is a terminal node then
        return the heuristic value of node
    if maximizingPlayer then
        value := −∞
        for each child of node do
            value := max(value, alphabeta(child, depth − 1, α, β, FALSE))
            α := max(α, value)
            if α ≥ β then
                break (* β cut-off *)
        return value
    else
        value := +∞
        for each child of node do
            value := min(value, alphabeta(child, depth − 1, α, β, TRUE))
            β := min(β, value)
            if α ≥ β then
                break (* α cut-off *)
        return value

Setup

• I like to use python virtualenvwrapper you can look my repo to learn how to use it: https://github.com/feyil/Virtualenv-Virtualenvwrapper-Usage

$ mkvirtualenv ai-chess -p python3
$ workon ai-chess
(ai-chess)$ pip install PyQt5
(ai-chess)$ pip install python-chess
(ai-chess)$ python main.py

• You can also use provided requirements.txt file after you set up the environment for the pip installs.

(ai-chess)$ pip install -r requirements.txt

• You can change the initial board used inside of the main.py file

#initial board is created here
board = chess.Board("rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1")