2220 articles – 2619 references  [version française]
 HAL: hal-00577733, version 1
 arXiv: 1103.3512
 Wavelet penalized likelihood estimation in generalized partially linear models
 (2011-03-17)
 The paper deals with a semiparametric generalized partially linear regression model with unknown regression coefficients and an unknown nonparametric function. We present a maximum penalized likelihood procedure to estimate the components of the partial linear model introducing penalty based wavelet estimators. Asymptotic rates of the estimates of both the parametric and the nonparametric part of the model are given and quasi-minimax optimality is obtained under usual conditions in literature. We establish in particular that the $\ell^1$-penalty leads to an adaptive estimation. An algorithm is also proposed for implementation and simulations are used to illustrate the results.
 1: Institut Camille Jordan (ICJ) CNRS : UMR5208 – Université Claude Bernard - Lyon I – Ecole Centrale de Lyon – Institut National des Sciences Appliquées (INSA) - Lyon
 Subject : Mathematics/StatisticsStatistics/Statistics Theory
 Keyword(s): semiparametric models – generalized regression – generalized partially linear models – M-estimation – penalized loglikelihood estimation – wavelets – backfitting
Attached file list to this document:
 PDF
 gannaz_2011.pdf(609.1 KB)
 PS
 gannaz_2011.ps(2.8 MB)
 hal-00577733, version 1 http://hal.archives-ouvertes.fr/hal-00577733 oai:hal.archives-ouvertes.fr:hal-00577733 From: Irène Gannaz <> Submitted on: Thursday, 17 March 2011 11:30:06 Updated on: Thursday, 17 March 2011 21:32:21