Balanced Neurofuzzy Models. Oleg Mytnyk

Abstract. This paper is devoted to the problem of a high complexity of fuzzy knowledge bases which contain enor- mous number of compound fuzzy rules. In order to significantly decrease the number of fuzzy rules and increase their transparency we present balanced neurofuzzy models. These models use the idea of Gabor-Kolmogorov expansion for ad- ditive decomposition into univariate and bivariate neurofuzzy submodels as well as maximum entropy principle to ground independent use of these submodels. Each submodel generates simplified rules independently of other submodels and contributes to fuzzy knowledge base of reduced complexity. The last but not least advantage of balanced neurofuzzy mod- els is that they can be regularized and learned by modern inductive methods. Although the present paper omits learning. We demonstrate the potential of balanced neurofuzzy approach on a toy example of wind-induced wave model.

Keywords. Fuzzy knowledge base (FKB), neurofuzzy model.

References.

1. Castellano G.: A neurofuzzy methodology for predictive modeling. Ph.D. thesis. Department of Computer Science, University of Bari, Italy, 2001.

2. Brown M., Harris C. J.: Neurofuzzy adaptive modelling and control. Hemel Hempstead, Prentice Hall, 1994.

3. Hong X., Harris C. J., Chen S.: Robust neurofuzzy rule base knowledge extraction and estimation using subspace de- composition combined with regularization and D-optimality. IEEE Transactions on Systems, Man, and Cybernetics, Part B, SMC-34(1): 598-608, 2004.

4. Roubos H., Setnes M.: Compact fuzzy models through complexity reduction and evolutionary optimization.In Proc. of the Ninth IEEE International Conference on Fuzzy Systems, vol. 2: 762-767, San Antonio, Texas, May 2000.

5. Hong X., Harris C. J.: Generalized neurofuzzy network modeling algorithms using Bezier-Bernstein polynomial functions and additive decomposition. IEEE Transactions On Neural Networks, 11(4): 889–902, 2000.

6. Jaynes E. T.: Prior probabilities.IEEE Transactions On Systems Science and Cybernetics, SSC-4(3): 227-241, 1968.

7. Jaynes E. T.: Probability Theory: The Logic of Science. Cambridge University Press, 2003.

8. Mytnik O. Yu.: Construction of Bayesian support vector regression in the feature space spanned by Bezier-Bernstein polynomial functions. Cybernetics and Systems Analysis, 43(4): 613-620, 2007.