Local Move Prediction in Go
The paper presents a system that learns to predict local strong expert moves in the game of Go at a level comparable to that of strong human kyu players. This performance is achieved by four techniques. First, our training algorithm is based on a relative-target approach that avoids needless weight adaptations characteristic of most neural-network classifiers. Second, we reduce dimensionality through state-of-the-art feature extraction, and present two new feature-extraction methods, the Move Pair Analysis and the Modified Eigenspace Separation Transform. Third, informed pre-processing is used to reduce state-space complexity and to focus the feature extraction on important features. Fourth, we introduce and apply second-phase training, i.e., the retraining of the trained network with an augmented input constituting all pre-processed features. Experiments suggest that local move prediction will be a significant factor in enhancing the strength of Go programs.
KeywordsLinear Discriminant Analysis Local Move Legal Move Multi Layer Perceptron Scatter Matrix
Unable to display preview. Download preview PDF.
- 3.Enderton, H.: The Golem Go program. Technical Report CMU-CS-92-101, School of Computer Science, Carnegie-Mellon University (1991)Google Scholar
- 4.Dahl, F.: Honte, a Go-playing program using neural nets. In: 16th International Conference on Machine Learning (1999)Google Scholar
- 5.Schraudolph, N., Dayan, P., Sejnowski, T.: Temporal difference learning of position evaluation in the game of Go. In: Cowan, J., Tesauro, G., Alspector, J. (eds.) Advances in Neural Information Processing 6, pp. 817–824. Morgan Kaufmann, San Francisco (1994)Google Scholar
- 6.Tesauro, G.: Connectionist learning of expert preferences by comparison training. In: Touretzky, D. (ed.) Advances in Neural Information Processing Systems 1 (NIPS 1988), pp. 99–106. Morgan Kaufmann, San Francisco (1989)Google Scholar
- 7.Riedmiller, M., Braun, H.: A direct adaptive method for faster backpropagation: the RPROP algorithm. In: IEEE Int. Conf. on Neural Networks (ICNN), pp. 586–591 (1993)Google Scholar
- 8.Jain, A., Chandrasekaran, B.: Dimensionality and sample size considerations in pattern recognition practice. In: Krishnaiah, P., Kanal, L. (eds.) Handbook of Statistics, vol. 2, pp. 835–855. North-Holland, Amsterdam (1982)Google Scholar
- 10.Bishop, C.: Neural Networks for Pattern Recognition. Clarendon Press, Oxford (1995)Google Scholar
- 11.Jollife, I.: Principal Component Analysis. Springer, Heidelberg (1986)Google Scholar
- 14.Kohonen, T.: Self-organising maps. Springer, Heidelberg (1995)Google Scholar
- 16.van der Werf, E.: Non-linear target based feature extraction by diabolo networks. Master’s thesis, Pattern Recognition Group, Department of Applied Physics, Faculty of Applied Sciences, Delft University of Technology (1999)Google Scholar
- 18.Müller, M.: Computer Go as a sum of local games: An application of combinatorial game theory. PhD thesis, ETH Zürich (1995) Diss. ETH No. 11.006Google Scholar