Fuzzy hierarchical cross-classification of Greek muds

In this paper we analyze the set of Greek muds from eight different locations given in ref 16 using a divisive fuzzy hierarchical cross-classification algorithm. We consider the fuzzy clustering algorithms are capable to eliminate the disfunctionalities of the hard clustering algorithms as well as to provide information obtained from a metrical analysis of the data. The fuzzy hierarchical cross-classification algorithm presented here produces not only a fuzzy partition of the muds in discussion but also a fuzzy partition of the 23 chemical and mineralogical characteristics, so that to each class of muds we may associate the class of characteristics that contributed to the separation of the class of muds.