Additive spectral method for fuzzy cluster analysis of similarity data including community structure and affinity matrices

An additive spectral method for fuzzy clustering is proposed. The method operates on a clustering model which is an extension of the spectral decomposition of a square matrix. The computation proceeds by extracting clusters one by one which makes the spectral approach quite natural. The iterative extraction of clusters, also, allows us to draw several stopping rules to the procedure. This applies to several relational data types differently normalized: network structure data (the first eigen-vector subtracted), affinity between multidimensional vectors (the pseudo-inverse Laplacian transformation), and conventional relational data including in-house data of similarity between research topics according to working of a research center. We experimentally compare the performance of our method with that of several recent techniques and show its competitiveness. Preprint submitted to Elsevier April 26, 2011