A survey of fuzzy clustering algorithms for pattern recognition - Part II

In Part I of this paper [1], an equivalence between the concepts of fuzzy c lustering and soft competitive learning in clustering algorithms is propose d on the basis of the existing literature. Moreover, a set of functional at tributes is selected for use as dictionary entries in the comparison of clu stering algorithms. In this paper, five clustering algorithms taken from the literature are rev iewed, assessed and compared on the basis of the selected properties of int erest. These clustering models are 1) self-organizing map (SOM); 2) fuzzy learning vector quantization (FLVQ); 3) fuzzy adaptive resonance theory (fuzzy ART); 4) growing neural gas (GNG); 5) fully self-organizing simplified adaptive resonance theory (FOSART). Although our theoretical comparison is fairly simple, it yields observation s that may appear paradoxical. First, only FLVQ, fuzzy ART, and FOSART expl oit concepts derived from fuzzy set theory (e.g., relative and/or absolute fuzzy membership functions). Secondly, only SORI, FLVQ, GNG, and FOSART emp loy soft competitive learning mechanisms, which are affected by asymptotic misbehaviors in the case of FLVQ, i.e., only SOM, GNG, and FOSART are consi dered effective fuzzy clustering algorithms.