Fiche du document
2016
- handle: 10670/1.00u2og
- hal: hal-03392022
- doi: 10.1214/15-AOS1376
- SCIENCESPO: 2441/etefo8s8r89oamhnhiclqr530
Ce document est lié à :
info:eu-repo/semantics/altIdentifier/doi/10.1214/15-AOS1376
Ce document est lié à :
info:eu-repo/semantics/altIdentifier/hdl/2441/etefo8s8r89oamhnhiclqr530
Ce document est lié à :
info:eu-repo/grantAgreement/EC/FP7/269693/EU/Wage Dynamics, Sorting Patterns in Labour Markets and Policy Evaluation/WASP
info:eu-repo/semantics/OpenAccess
Mots-clés
Estimators Markov models Diagonal argumentsSujets proches
Field guides Keys (Identification guides) Identification guides Forensic identificationCiter ce document
Stéphane Bonhomme et al., « Estimating Multivariate Latent-Structure Models », Archive ouverte de Sciences Po (SPIRE), ID : 10.1214/15-AOS1376
Métriques
Partage / Export
Résumé
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 joint-diagonalization problem. Algorithms are available for computation and we derive distribution theory. We further develop asymptotic theory for orthogonal-series estimators of component densities in mixture models and emission densities in hidden Markov models.