David R Newman

Third Year Project

A Game of Intelligent Risk

A copy of my final project report can be found here.

Project Abstract

This project attempts to determine whether the rules of the well-known board game "Risk" can be adapted to generate a game model that then permits the development of Artificial Intelligence (AI) players that can play the game. This game was chosen as it provides an environment that allows analysis of some of central processes of AI.

The development game model comprised of several parts, a verification engine, a Graphical User Interface (GUI) to interact with the game model and a general infrastructure to support the AI players, which includes methods to automate and benchmark them.

Project Implementation & Source Code

All the source code and compiled class files can be found in source_code.zip. The application works perfectly fine on WindowsXP Pro running Java1.4. Obviously as I have not needed to use this application for some time I cannot gurantee how long you will have to play around with the source to get it to work on other OS and Java Versions.

To run the application in GUI mode type at the command line:

java CustomPanelTest

To run in stats collection mode type:

java CustomPanelTest <no of games> <rules file> <positions file> 
  <stats output file> <tournament file> <reinforcement-learning file>

substituting the appropriate filenames in. For further information refer to my final report.

Page written by David R Newman (drn[at]ecs.soton.ac.uk). Last updated October 13 2009 18:12:07.