|
Abstract:
|
The natural immune system found in mammals offers interesting insights for models with features such as decentralized control, pattern recognition and learning. The work presented here experiments with the effectiveness of using the natural immune system model as a method for discovering heuristics in the game of Othello. The features of the AIS that offer a novel approach to decision making in this context are compared to the classic heuristics of "lookahead", where an algorithm considers all moves in the game to find the best move. The game of Othello is used since it offers a compelling search space and simple rules. Claims as to the appropriateness of using the natural immune system paradigm for discovering heuristics in Othello, and what the results may imply for other games both like and unlike Othello, will be discussed.
|
