Skip to Main content Skip to Navigation

Sélection de modèles semi-paramétriques

Abstract : This thesis develops model selection method for Biostatistical applications, essentially in medicine. The study is structured into three parts. In the first part, we propose a method and a program to determine a significance level for a series of codings of an explanatory variable in logistic regression. This method is illustrated using the data of a study of the relation between cholesterol and dementia. The second part of the thesis presents a general criterion for choosing an estimator in a family of semi-parametric estimators. It is based on a bootstrap estimator of the Kullback-Leibler information. This criterion is applied for modeling the effect of the asbestos on the risk of mesotheliome and we compare it to Birge-Massart selection method. The last part provides a criterion for choosing an estimator in a family of semi-parametric estimators from incomplete data. It is an extension of the criterion presented in the second part. This criterion is based on the expected observed log-likelihood. It is used in particular to select the smoothing parameter for the hazard function and to choose between stratified and unstratified proportional hazards models. An example is given for modeling the effect of sex and educational level on the risk of developing dementia.
Document type :
Complete list of metadata
Contributor : Benoit Liquet <>
Submitted on : Friday, February 21, 2003 - 1:09:47 PM
Last modification on : Tuesday, April 28, 2020 - 1:02:45 AM
Long-term archiving on: : Tuesday, September 11, 2012 - 7:55:16 PM


  • HAL Id : tel-00002430, version 1



Benoit Liquet. Sélection de modèles semi-paramétriques. Mathématiques [math]. Université Victor Segalen - Bordeaux II, 2002. Français. ⟨tel-00002430⟩



Record views


Files downloads