Minimax search and Alpha-beta pruning for game playing
We develop a Minimax and alpha-beta search based intelligent agent for Warfare game. We implement the above with GUI using Qt
Installation
- Install Qt-C++ and the dependant libraries.
- Download the game and Compile the game in Release Mode.
Tested and verified in Qt 5.3 on Windows
Game Play
Select the board, agent for player 1 and player 2. If any of the player is an automated agent (Minimax or A-B pruning), a difficulty slider will be displayed. Set the required difficulty level (1-5). If both the agents are A-B pruning,the difficulty can be set upto level 7, but note that game play is slower for higher difficulty. Default values are already set ( Board-Keren, Player1-Human, Player2-Human, Difficulty level-3). Click on Start Game to begin. If one or more players are Human, the game can be played by clicking the buttons on the grid. There is a status label which displays whose turn it is or the status (Begin Game/Game over along with the winner). The LCD score boards show the current scores at every instant during game play.
Game Logs
Additionally after every game, we store the game data in a text file at the location ./build-Warfare-Desktop_Qt_5_3_MSVC2013_OpenGL_64bit-Release.
Read ./Outputs/ReadMeforGameData.txt for further details on how to find the file and interpret the data. There is a RunAllCombo button which automatically plays the game for all combinations of board, depth and and algorithm for automated agents and logs the information.
Others
Project submitted as part of Artificial Intelligence Course (CS60045), IIT Kharagpur. Check out the report as well.