Estimating Multivariate Latent-Structure Models - Sciences Po Access content directly
Preprints, Working Papers, ... Year : 2014

Estimating Multivariate Latent-Structure Models

Abstract

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.
Fichier principal
Vignette du fichier
2014-18.pdf (536.98 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-01097135 , version 1 (22-12-2014)
hal-01097135 , version 2 (20-08-2015)

Identifiers

Cite

Jean-Marc Robin, Stéphane Bonhomme, Koen Jochmans. Estimating Multivariate Latent-Structure Models. 2014. ⟨hal-01097135v1⟩

Collections

SCPO-ECON-DP
181 View
298 Download

Share

Gmail Mastodon Facebook X LinkedIn More