Evolving a Neural Network to Play Checkers without Human Expertise
We have been exploring the potential for a co-evolutionary process to learn how to play checkers without relying on the usual inclusion of human expertise in the form of features that are believed to be important to playing well. In particular, we have focused on the use of a population of neural networks, where each network serves as an evaluation function to describe the quality of the current board position. After only a little more than 800 generations, the evolutionary process has generated a neural network that can play checkers at the expert level as designated by the U.S. Chess Federation rating system. This has been documented against real players with games played over the Internet. Our checkers program, named Anaconda, has also competed well against commercially available software.
KeywordsNeural Network Hide Layer Evolutionary Computation Hide Node Radial Basis Function Network
Unable to display preview. Download preview PDF.
- Schaeffer, J. (1996), One Jump Ahead: Challenging Human Supremacy in Checkers, Springer, Berlin.Google Scholar
- Chellapilla. K. and Fogel, D. B. (2000) “Evolving an expert checkers playing program without using human expertise,” IEEE Trans. Pattern Analysis and Machine Intelligence, in review.Google Scholar
- Yao, X. and Fogel, D. B. (2000), Proceedings of the First IEEE Symposium on Combinations of Evolutionary Computation and Neural Networks,IEEE Press, Piscataway, NJ, in press.Google Scholar