Comparative Analysis of Computational Intelligence Techniques in Chess Player Classification

• Pedro H. A. Ribeiro UFPI
• Ronnyel C. C. Silva UFPI
• José M. P. M. Júnior UFPI

Abstract

In this work, the Multi-Layer Perceptron (MLP), Radial Base Network (RBF), and Learning Vector Quantization (LVQ) neural patterns are repeated to solve the problem of classification of chess players in levels (basic, intermediate and advanced). The following parameters were used as input to neural networks as indicators of chess match performance: peer media, inaccuracies, errors and grave errors. For each technique used, four network topologies were used, which varied the number of hidden layer neurons. Performance measures were mean acuity, minimum acuity, minimum acuity, mean acuity, standard deviation of actions, and mean square error (MSE). The results were more than an MLP obtained as the best performance solutions in solving the chess player classification problem.

Keywords: Neural Networks, Chess Game, Player Rankings

References

Bishop, Christopher M. et al. Neural networks for pattern recognition. Oxford university press, 1995.

Bonifácio, F. N. Comparação entre as Redes Neurais Artificiais MLP, RBF e LVQ na Classificação de Dados. Paraná: Universidade Estadual do Oeste do Paraná, 2010.

Chollet, F. Deep Learning mit Python und Keras: Das Praxis-Handbuch vom Entwickler der Keras-Bibliothek. MITP-Verlags GmbH & Co. KG, 2018.D'Agostini, Orfeu Gilberto. Xadrez básico. Ediouro Publicações, 2002.

D'Agostini, Orfeu Gilberto. Xadrez básico. Ediouro Publicações, 2002.

Dauvergne, P. The case for chess as a tool to develop our children’s mind. University of Sidney. Disponível em: < http://www.auschess.org.aurticles/chessmind.htm.> Acesso em: 20 de março de 2019.

Erilli, N. A. et al. (2010) Classifying Chess Players with Fuzzy Clustering Analysis in Fuzzy Data Using Eco Codes. In: 1st International Symposium on Computing in Science & Engineering Proceedings Book. p. 1105-1110.

Fernandes, M. A. C. Aplicação das Redes RBF na Deteção Inteligente de Sinais Digitais. Proceeding of the IV Brazilian Conference on Neural Networks. Pp 226-230. São José dos Campos. São Paulo. Brazil.

HU, Yu Hen; HWANG, Jenq-Neng. Introduction to neural networks for signal processing. Handbook of neural network signal processing, p. 12-41, 2002.

Jeremic, V. M.; Radojicic, Z. (2010) A new approach in the evaluation of team chess championships rankings. Journal of Quantitative Analysis in Sports, v. 6, n. 3.

Kohonen, T. K. (1997) Self-Organizing Maps. Springer-Verlag, Berlin, Heidelberg, 2nd extended edition.

Kugler, M., Júnior, J.T., Lopes, H. S. (2003) Desenvolvimento de uma Rede Neural LVQ em Linguagem VHDL para Aplicações em Tempo-Real. Proceedings of the VI Brazilian Conference on Neural Networks - VI Congresso Brasileiro de Redes Neurais pp. 103–108, June 2–5, 2003 – Centro Universitário da FEI, São Paulo - SP – Brazil.

Reingold, E. M. et al. (2001) Visual span in expert chess players: Evidence from eye movements. Psychological Science, v. 12, n. 1, p. 48-55.

Santos, A. B., Castanho, M. J. P., Hernandes, F., Ré, A. M. (2010) Rede Neural LVQ em Câncer de Próstata. Anais do XXXIII Congresso Nacional de Matemática Aplicada e Computacional.

Silva, IN da; SPATTI, Danilo H.; Flauzino, R. A. Redes neurais artificiais para engenharia e ciências aplicadas. São Paulo: Artliber, v. 23, n. 5, p. 33-111, 2010.

Silva, W. (2010) Raciocínio logico e o jogo de xadrez = em busca de relações. 2 v. Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Educação, Campinas, SP. Disponível em: . Acesso em: 20 de março de 2019.

Poggio, T., Girosi, T. (1990) Networks for approximation and learning. Proceedings of the IEEE, 78:1481–1497.