Parameter Setting for Evolutionary Latent Class Clustering
Damien Tessier, Marc Schoenauer, Christophe Biernacki, Gilles Celeux and Gérard Govaert
In: Second International Symposium, ISICA 2007, 21-24 Sept 2007, Wuhan, China.
The latent class model or multivariate multinomial mixture is a powerful model for clustering discrete data. This model is expected to be useful to represent non-homogeneous populations. It uses a conditional independence assumption given the latent class to which a statistical unit is belonging. However, it leads to a criterion that proves difficult to optimise by the standard approach based on the EM algorithm.
An Evolutionary Algorithms is designed to tackle this discrete
optimisation problem, and an extensive parameter study on a large artificial dataset allows to derive stable parameters. A Monte Carlo approach is used to validate those parameters on other artificial datasets, as well as on some well-known real data: the Evolutionary Algorithm seems to repeatedly perform better than other standard clustering techniques on the same data.