Open AccessVariable selection for a mark-specific additive hazards model using the adaptive LASSO
Author(s) -
Dongxiao Han,
Lianqiang Qu,
Liuquan Sun,
Yanqing Sun
Publication year - 2021
Publication title -
statistical methods in medical research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.952
H-Index - 85
eISSN - 1477-0334
pISSN - 0962-2802
DOI - 10.1177/09622802211023957
Subject(s) - lasso (programming language) , covariate , estimator , feature selection , proportional hazards model , sample size determination , selection (genetic algorithm) , computer science , econometrics , variable (mathematics) , model selection , vaccine efficacy , mathematics , statistics , medicine , artificial intelligence , vaccination , mathematical analysis , world wide web , immunology
In HIV vaccine efficacy trials, mark-specific hazards models have important applications and can be used to evaluate the strain-specific vaccine efficacy. Additive hazards models have been widely used in practice, especially when continuous covariates are present. In this article, we conduct variable selection for a mark-specific additive hazards model. The proposed method is based on an estimating equation with the first derivative of the adaptive LASSO penalty function. The asymptotic properties of the resulting estimators are established. The finite sample behavior of the proposed estimators is evaluated through simulation studies, and an application to a dataset from the first HIV vaccine efficacy trial is provided.