Some clustering techniques for modelling uncertain nonlinear systems

Abstract

This paper presents popular unsupervised clustering algorithms based on neuro-fuzzy, fuzzy c-means (FCM) and agglomerative techniques. The purpose of this paper is to provide clustering methods able to cluster the data patterns without a priori information about the number of clusters. We will show that it is possible to reconcile the FCM algorithm with the unsupervised clustering algorithms. Finally, to show the efficiencies of these algorithms, we will apply them to model the behaviour of uncertain system.