Skip to Main content Skip to Navigation
Conference papers

Intégrer les données manquantes dans la sélection de variables pour données longitudinales

Abstract : Generalized estimating equations (GEE) are a useful tool for marginal regression analysis with repeated measurements. Missing data as well as a large number of variables combined with small sample size are usual issues faced with longitudinal data. Multiple imputation is a popular tool for handling missing data and in particular , the MI-GEE can be used for inference. The multiple imputation-least absolute shrinkage and selection operator (MI-LASSO) proposes a consistent selection through the multiply-imputed datasets but cannot handle correlation among individual observations. 1 We present MI-PGEE, a new multiple imputation-penalized generalized estimating equations as an extension of the MI-LASSO to be applied on longitudinal data. MI-PGEE applies the penalized GEE with ridge penalty and adaptive weights that are common to the group of estimated regression coefficients of the same variable across multiply-imputed datasets. In order to select the tuning parameter, a new BIC-like criterion is presented. MI-PGEE yields a consistent variable selection across multiply-imputed datasets, making this a selection method for longitudinal data able to manage missing data and within subject correlation. The usefulness of the new method is illustrated by an application on the placebo arm of the Strontium ranelate Efficacy in Knee OsteoarthrItis triAl (SEKOIA) study.
Document type :
Conference papers
Complete list of metadatas

Cited literature [10 references]  Display  Hide  Download

https://hal-cnam.archives-ouvertes.fr/hal-02500612
Contributor : Gilbert Saporta <>
Submitted on : Friday, March 6, 2020 - 11:04:33 AM
Last modification on : Thursday, March 12, 2020 - 1:24:56 AM
Document(s) archivé(s) le : Sunday, June 7, 2020 - 12:56:23 PM

File

geronimi_saporta_Jds2016.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02500612, version 1

Collections

Citation

Julia Geronimi, Gilbert Saporta. Intégrer les données manquantes dans la sélection de variables pour données longitudinales. 48 èmes Journées de Statistique, May 2016, Montpellier, France. ⟨hal-02500612⟩

Share

Metrics

Record views

9

Files downloads

9