A Copula-based clustering algorithm

In this work we investigate the problem of clustering dependent data by means of copula functions. We propose a new algorithm (CoClust) that allows to cluster dependent data according to the multivariate structure of the generating process without any assumptions on the margins. Our approach does not require either to choose a starting classification or to set a priori the exact number of clusters. We discuss some theoretical issues and test our proposal on simulated data for different dependence scenarios by comparing it with other model–based clustering techniques. Finally, we show applications to microarray data by using the R package CoClust that we have developed on purpose (see http://www2.stat.unibo.it/giannerini/software.html).