Mixture of Gaussian regressions model with logistic weights, a penalized maximum likelihood approach

Lucie Montuelle; Erwan Le Pennec

doi:10.1214/14-EJS939

Mixture of Gaussian regressions model with logistic weights, a penalized maximum likelihood approach

Lucie Montuelle ^{1, 2} Erwan Le Pennec ^{3, 2}

1 LM-Orsay - Laboratoire de Mathématiques d'Orsay

2 SELECT - Model selection in statistical learning

Inria Saclay - Ile de France, LMO - Laboratoire de Mathématiques d'Orsay, CNRS - Centre National de la Recherche Scientifique : UMR

3 X - École polytechnique

Abstract : In the framework of conditional density estimation, we use candidates taking the form of mixtures of Gaussian regressions with logistic weights and means depending on the covariate. We aim at estimating the number of components of this mixture, as well as the other parameters, by a penalized maximum likelihood approach. We provide a lower bound on the penalty that ensures an oracle inequality for our estimator. We perform some numerical experiments that support our theoretical analysis.

Keywords : Mixture of Gaussian regressions models mixture of regressions models penalized likelihood model selection

Document type :

Journal articles

Domain :

Mathematics [math] / Statistics [math.ST]

Liste complète des métadonnées

https://hal.inria.fr/hal-01101483
Contributor : Lucie Montuelle <>
Submitted on : Thursday, January 8, 2015 - 5:08:04 PM
Last modification on : Thursday, February 7, 2019 - 2:25:28 PM

Mixture of Gaussian regressions model with logistic weights, a penalized maximum likelihood approach

Links full text

Identifiers

Collections

Citation

Export

Metrics

243

Contact

Mixture of Gaussian regressions model with logistic weights, a penalized maximum likelihood approach

Links full text

Identifiers

Collections

Citation

Export

Share

Metrics

243

Contact