Multiple imputation with compatibility for high-dimensional data.

Multiple Imputation (MI) is always challenging in high dimensional settings. The imputation model with some selected number of predictors can be incompatible with the analysis model leading to inconsistent and biased estimates. Although compatibility in such cases may not be achieved, but one can ob...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Faisal Maqbool Zahid, Shahla Faisal, Christian Heumann
Formato:	article
Lenguaje:	EN
Publicado:	Public Library of Science (PLoS) 2021
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/cc55db66f7de4a9db934f51c1ba44619
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:cc55db66f7de4a9db934f51c1ba44619
record_format	dspace
spelling	oai:doaj.org-article:cc55db66f7de4a9db934f51c1ba446192021-12-02T20:19:40ZMultiple imputation with compatibility for high-dimensional data.1932-620310.1371/journal.pone.0254112https://doaj.org/article/cc55db66f7de4a9db934f51c1ba446192021-01-01T00:00:00Zhttps://doi.org/10.1371/journal.pone.0254112https://doaj.org/toc/1932-6203Multiple Imputation (MI) is always challenging in high dimensional settings. The imputation model with some selected number of predictors can be incompatible with the analysis model leading to inconsistent and biased estimates. Although compatibility in such cases may not be achieved, but one can obtain consistent and unbiased estimates using a semi-compatible imputation model. We propose to relax the lasso penalty for selecting a large set of variables (at most n). The substantive model that also uses some formal variable selection procedure in high-dimensional structures is then expected to be nested in this imputation model. The resulting imputation model will be semi-compatible with high probability. The likelihood estimates can be unstable and can face the convergence issues as the number of variables becomes nearly as large as the sample size. To address these issues, we further propose to use a ridge penalty for obtaining the posterior distribution of the parameters based on the observed data. The proposed technique is compared with the standard MI software and MI techniques available for high-dimensional data in simulation studies and a real life dataset. Our results exhibit the superiority of the proposed approach to the existing MI approaches while addressing the compatibility issue.Faisal Maqbool ZahidShahla FaisalChristian HeumannPublic Library of Science (PLoS)articleMedicineRScienceQENPLoS ONE, Vol 16, Iss 7, p e0254112 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Faisal Maqbool Zahid Shahla Faisal Christian Heumann Multiple imputation with compatibility for high-dimensional data.
description	Multiple Imputation (MI) is always challenging in high dimensional settings. The imputation model with some selected number of predictors can be incompatible with the analysis model leading to inconsistent and biased estimates. Although compatibility in such cases may not be achieved, but one can obtain consistent and unbiased estimates using a semi-compatible imputation model. We propose to relax the lasso penalty for selecting a large set of variables (at most n). The substantive model that also uses some formal variable selection procedure in high-dimensional structures is then expected to be nested in this imputation model. The resulting imputation model will be semi-compatible with high probability. The likelihood estimates can be unstable and can face the convergence issues as the number of variables becomes nearly as large as the sample size. To address these issues, we further propose to use a ridge penalty for obtaining the posterior distribution of the parameters based on the observed data. The proposed technique is compared with the standard MI software and MI techniques available for high-dimensional data in simulation studies and a real life dataset. Our results exhibit the superiority of the proposed approach to the existing MI approaches while addressing the compatibility issue.
format	article
author	Faisal Maqbool Zahid Shahla Faisal Christian Heumann
author_facet	Faisal Maqbool Zahid Shahla Faisal Christian Heumann
author_sort	Faisal Maqbool Zahid
title	Multiple imputation with compatibility for high-dimensional data.
title_short	Multiple imputation with compatibility for high-dimensional data.
title_full	Multiple imputation with compatibility for high-dimensional data.
title_fullStr	Multiple imputation with compatibility for high-dimensional data.
title_full_unstemmed	Multiple imputation with compatibility for high-dimensional data.
title_sort	multiple imputation with compatibility for high-dimensional data.
publisher	Public Library of Science (PLoS)
publishDate	2021
url	https://doaj.org/article/cc55db66f7de4a9db934f51c1ba44619
work_keys_str_mv	AT faisalmaqboolzahid multipleimputationwithcompatibilityforhighdimensionaldata AT shahlafaisal multipleimputationwithcompatibilityforhighdimensionaldata AT christianheumann multipleimputationwithcompatibilityforhighdimensionaldata
_version_	1718374192307503104

Multiple imputation with compatibility for high-dimensional data.

Ejemplares similares