Events and Meetings of Italian Statistical Society, Advances in Latent Variables - Methods, Models and Applications

Font Size: 
Composite likelihood inference for a class of latent variable models for two factor clustering
Francesco Bartolucci, Francesca Chiaromonte, Prabhani Kuruppumullage, Bruce G. Lindsay

Last modified: 2013-06-14

Abstract


We consider a discrete latent variable model for arrays of data, which allows for two factor clustering of the observed units when one dimension is referred to consecutive time occasions. The model then relies on a hidden Markov structure but, given its complexity, cannot be estimated by full maximum likelihood. Therefore, we introduce a composite likelihood approach based on considering different subsets of data. The proposed approach is illustrated by a simulation study and an application in genomics.

Full Text: PDF