Gaussian topographic co-clustering model

Abstract

The visualization of the clusters obtained by a partitioning procedure is very informative as this helps to a better overview of the contents of a data table. For co-clustering, the latent block mixture model is very effective. We propose to define generative self-organizing maps with this model for Gaussian blocks. A perspective is the analysis and the visualization of continuous data. © 2013 Springer-Verlag.