Teaching Introductory Artificial Intelligence through Java-based Games

Abstract

We introduce a Java graphical gaming framework that enables students in an introductory artificial intelligence (AI) course to immediately apply and visualize the topics from class. We have used this framework in teaching a mixed undergraduate/graduate AI course for six years. We believe that the use of games motivates students. The graphical nature of each game enables students to quickly see how well their algorithm works. Because the topics in an introductory AI course vary widely, students apply their algorithms to multiple game environments. A final challenging environment enables them to tie together the concepts for the entire semester.

Paper

McGovern, Amy and Tidwell, Zachery and Rushing, Derek (2011). Teaching Introductory Artificial Intelligence through Java-based Games. To appear in the proceedings of the symposium on Educational Advances in Artificial Intelligence. [pdf preprint (924 K)]

Code release

The code is available as a zip file here. It has all 4 games (spacewar, mancala, and pacman are all in one of the frameworks and then roborally in a very similar framework). The code is being released using the MIT Open Source License.

Example assignments

Getting started

• An example of a project 0 where the students check out the code, make a sample (random) agent, change their password, and turn in the code

Spacewar

• A* navigation (single agent environment, typical first assignment in spacewar)
• Learning using evolutionary computation only
• Learning using multiple methods (evolutionary computation plus a method of their choice)
• Capture the flag (multi-player teams, typical final assignment)

Pacman

• Breadth first search and depth first search
• A* search to avoid ghosts as well as another informed search method (such as hill climbing)
• Evolutionary computation to enable Pacman to move intelligently

Mancala

• Playing intelligently using minimax search with and without alpha-beta pruning
• Using learning to improve the evaluation function for minimax search

Roborally

• A* to move the robots around intelligently. Can also use rollouts.
• Evolutionary computation to have the robots learn, potentially using option cards
• Small teams of robots trying to capture the flag as a team (multi-agent coordination)