PARAMETRIC AND NONPARAMETRIC UNSUPERVISED CLUSTER-ANALYSIS

Much work has been published on methods for assessing the probable num ber of clusters or structures within unknown data sets. This paper aim s to look in more detail at two methods, a broad parametric method, ba sed around the assumption of Gaussian clusters and the other a non-par ametric method which utilises methods of scale-space filtering to extr act robust structures within a data set. It is shown that, whilst both methods are capable of determining cluster validity for data sets in which clusters tend towards a multivariate Gaussian distribution, the parametric method inevitably fails for clusters which have a non-Gauss ian structure whilst the scale-space method is more robust. Copyright (C) 1997 Pattern Recognition Society.