Auteurs Centres de recherche Disciplines et Collections Projets
Français English
 
 

Estimating Multivariate Latent-Structure Models

 

Notice

Type:   Working paper
 
Titre:   Estimating Multivariate Latent-Structure Models
 
Auteur(s):   Robin, Jean-Marc (1961-...) - Département d'économie (Auteur)
Bonhomme, Stéphane - University of Chicago (Auteur)
Jochmans, Koen (1982-...) - Département d'économie (Auteur)
 
Date de publication:   2014-12
 
Collection:   Sciences Po Economics Discussion Papers  :  2014-18
 
Mots-clés:   [en] hidden Markov model, finite mixture model, latent structure, multilinear restrictions, multivariate data, nonparametric estimation, simultaneous matrix diagonalization.
 
Résumé:   [en] A constructive proof of identification of multilinear decompositions of multiway arrays is presented. It can be applied to show identification in a variety of multivariate latent structures. Examples are finite-mixture models and hidden Markov models. The key step to show identification is the joint diagonalization of a set of matrices in the same non-orthogonal basis. An estimator of the latent-structure model may then be based on a sample version of this simultaneous-diagonalization problem. Simple algorithms are available for computation. Asymptotic theory is derived for this joint approximate-diagonalization estimator.
 
 

Fichiers

Version Format Fichier Taille
Version de l'auteur 2014-18.pdf 0,52 MB