Section: New Results

Fuzzy Clustering on Multiple Dissimilarity Matrices

Participants : Marc Csernel, Yves Lechevallier.

This work is done in collaboration with F.A.T de Carvalho (University of Recife, Brazil) [41] .

The goal of the fuzzy clustering algorithms is to partition objects taking into account simultaneously their relational descriptions given by multiple dissimilarity matrices. The aim is to obtain a collaborative role of the different dissimilarity matrices in order to obtain a final consensus partition. These matrices could have been obtained using different sets of variables and dissimilarity functions. These algorithms are designed to furnish a partition and a prototype for each fuzzy cluster as well as to learn a relevance weight for each dissimilarity matrix by optimizing an adequacy criterion that measures the fitting between the fuzzy clusters and their representatives.

These relevance weights change at each algorithm iteration and can either be the same for all fuzzy clusters or different from one fuzzy cluster to another. Experiments with real-valued datasets from UCI machine learning repository as well as symbolic data sets show the usefulness of the proposed fuzzy clustering algorithms.