A mixture of local and quadratic approximation variable selection algorithm in nonconcave penalized regression - Archive ouverte HAL Access content directly
Journal Articles Revue Africaine de Recherche en Informatique et Mathématiques Appliquées Year : 2013

A mixture of local and quadratic approximation variable selection algorithm in nonconcave penalized regression

Abstract

We consider the problem of variable selection via penalized likelihood using nonconvex penalty functions. To maximize the non-differentiable and nonconcave objective function, an algorithm based on local linear approximation and which adopts a naturally sparse representation was recently proposed. However, although it has promising theoretical properties, it inherits some drawbacks of Lasso in high dimensional setting. To overcome these drawbacks, we propose an algorithm (MLLQA) for maximizing the penalized likelihood for a large class of nonconvex penalty functions. The convergence property of MLLQA and oracle property of one-step MLLQA estimator are established. Some simulations and application to a real data set are also presented.
Fichier principal
Vignette du fichier
Vol.16.pp.29-46.pdf (862.41 Ko) Télécharger le fichier
Origin : Publisher files allowed on an open archive
Loading...

Dates and versions

hal-01299521 , version 1 (07-04-2016)

Identifiers

Cite

Assi N'Guessan, Ibrahim Sidi Zakari, Assi Mkhadri. A mixture of local and quadratic approximation variable selection algorithm in nonconcave penalized regression. Revue Africaine de Recherche en Informatique et Mathématiques Appliquées, 2013, Volume 16, 2012, pp.29-46. ⟨10.46298/arima.1962⟩. ⟨hal-01299521⟩

Collections

INSMI
74 View
792 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More