Applying radial basis function neural network for comprehending properties of each cluster of fuzzy c-means in coordinates analysis (case study in Iran)

Abstract

Abstract For notifying the properties of special area with similar properties, clustering analysis is really helpful, and neural network methods have ability to create usable model. One of the best ways for clustering is fuzzy c-means, and fuzzy c-means by the basis of fuzzy method divides data set to different clusters. Radial basis function is neural network which is utilizing spread and this algorithm’s layers like input layer, hidden layer and output layer for creating effective neural network. This paper is introduced a novel method, in this method data points (longitude and latitude of main cities of Iran) by using fuzzy c-mean algorithm is divided to different clusters then for each cluster RBF neural networks is defined separately, and this method is FCM-RBF. The outcome of FCM-RBF build neural network for each cluster separately, and result of this study shows that radial basis function neural network can enhance the quality of analysis of outcomes of this kind of clustering and by applying this algorithms different clusters with same properties is calculated and create neural network separately for each cluster, and three clusters are proposed for this algorithms and data points of cluster2 and cluster3 has acceptable rate of adaptability with RBF neural network but data points of cluster1 can’t adapt themselves with neural network perfectly, and validity of outcomes of this clustering increase by using radial basis function neural network. In this algorithm data points of each clusters can separately analyze which is cause better comprehending of study area.