LJKProbability & Statistics Seminar

On Thursday December 6 2012 at 14h00 in Salle 1  Tour IRMA

Seminary of Charles BOUVEYRON (Laboratoire SAMM, Université Paris 1)

Sparse and discriminative clustering for highdimensional data

Summary

The FisherEM algorithm has been recently proposed for the simultaneous visualization and clustering of highdimensional data. It is based on a mixture model which fits the data into a latent discriminative subspace with a low intrinsic dimension. From a practical point of view, the FisherEM algorithm turns out to outperform other subspace clustering in most situations. The convergence of the FisherEM algorithm is as well studied. It is in particular proved that the algorithm converges under weak conditions in the general case. It is also shown that the Fisher's criterion can be used as stopping criterion for the algorithm to improve the clustering accuracy and that the FisherEM algorithm usually converges faster than both the EM and CEM algorithms. Finally, a sparse extension of the FisherEM algorithm is proposed by adding a L1 constraint in the F step. This allows in particular to perform a selection of the original variables which are discriminative.
