Design of Optimized Radial Basis Function Neural Networks Classifier with the Aid of Fuzzy Clustering and Data Preprocessing Method
In this paper, we introduce a new architecture of optimized RBF neural network classifier with the aid of fuzzy clustering and data preprocessing method and discuss its comprehensive design methodology. As the pre-processing part, LDA algorithm is combined in front of input layer and then the new feature samples obtained through LDA are to be the input data of FRBF neural networks. In the hidden layer, FCM algorithm is used as receptive field instead of Gaussian function. The connection weights of the proposed model are used as polynomial function. PSO algorithm is also used to improve the accuracy and architecture of classifier. The feature vector of LDA, the fuzzification coefficient of FCM, and the polynomial type of RBF neural networks are optimized by means of PSO. The performance of the proposed classifier is illustrated with several benchmarking data sets and is compared with other classifier reported in the previous studies.
KeywordsRadial basis function neural network Fuzzy C-means clustering Particle swarm optimization Linear discriminant analysis
Unable to display preview. Download preview PDF.
- 1.Bishop, C.M.: Neural networks for pattern recognition. Oxford Univ., New York (1981)Google Scholar
- 4.McLachlan, G.J.: Discriminant Analysis and Statistical Pattern Recognition. Wiley Interscience (2004)Google Scholar
- 6.Duda, R.O., Hart, P.E., Stork, D.G.: Pattern classification, 2nd edn. Wiley-Intersicence (2000)Google Scholar
- 7.Kennedy, J., Eberhart, R.: Particle swarm optimization. In: Proc. IEEE Int. Conf. Neural Networks, vol. 4, pp. 1942–1948 (1995)Google Scholar
- 8.Duda, R.O., Hart, P.E.: Pattern classification and scene analysis. Wiley, New York (2002)Google Scholar