Linear and nonlinear variable selection in competing risks data

Xiaowei Ren, Shanshan Li, Changyu Shen, Zhangsheng Yu

Research output: Contribution to journalArticle

Abstract

Subdistribution hazard model for competing risks data has been applied extensively in clinical researches. Variable selection methods of linear effects for competing risks data have been studied in the past decade. There is no existing work on selection of potential nonlinear effects for subdistribution hazard model. We propose a two-stage procedure to select the linear and nonlinear covariate(s) simultaneously and estimate the selected covariate effect(s). We use spectral decomposition approach to distinguish the linear and nonlinear parts of each covariate and adaptive LASSO to select each of the 2 components. Extensive numerical studies are conducted to demonstrate that the proposed procedure can achieve good selection accuracy in the first stage and small estimation biases in the second stage. The proposed method is applied to analyze a cardiovascular disease data set with competing death causes.

Original languageEnglish (US)
JournalStatistics in Medicine
DOIs
StateAccepted/In press - Jan 1 2018
Externally publishedYes

Fingerprint

Competing Risks
Variable Selection
Covariates
Hazard Models
Proportional Hazards Models
Adaptive Lasso
Two-stage Procedure
Spectral Decomposition
Nonlinear Effects
Numerical Study
Cause of Death
Cardiovascular Diseases
Estimate
Demonstrate
Research

Keywords

  • Cubic b-spline
  • Penalized log-likelihood
  • Spectral decomposition
  • Subdistribution hazard

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability

Cite this

Linear and nonlinear variable selection in competing risks data. / Ren, Xiaowei; Li, Shanshan; Shen, Changyu; Yu, Zhangsheng.

In: Statistics in Medicine, 01.01.2018.

Research output: Contribution to journalArticle

@article{b718c7dfee484e6fbc85510935acd2c6,
title = "Linear and nonlinear variable selection in competing risks data",
abstract = "Subdistribution hazard model for competing risks data has been applied extensively in clinical researches. Variable selection methods of linear effects for competing risks data have been studied in the past decade. There is no existing work on selection of potential nonlinear effects for subdistribution hazard model. We propose a two-stage procedure to select the linear and nonlinear covariate(s) simultaneously and estimate the selected covariate effect(s). We use spectral decomposition approach to distinguish the linear and nonlinear parts of each covariate and adaptive LASSO to select each of the 2 components. Extensive numerical studies are conducted to demonstrate that the proposed procedure can achieve good selection accuracy in the first stage and small estimation biases in the second stage. The proposed method is applied to analyze a cardiovascular disease data set with competing death causes.",
keywords = "Cubic b-spline, Penalized log-likelihood, Spectral decomposition, Subdistribution hazard",
author = "Xiaowei Ren and Shanshan Li and Changyu Shen and Zhangsheng Yu",
year = "2018",
month = "1",
day = "1",
doi = "10.1002/sim.7637",
language = "English (US)",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",

}

TY - JOUR

T1 - Linear and nonlinear variable selection in competing risks data

AU - Ren, Xiaowei

AU - Li, Shanshan

AU - Shen, Changyu

AU - Yu, Zhangsheng

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Subdistribution hazard model for competing risks data has been applied extensively in clinical researches. Variable selection methods of linear effects for competing risks data have been studied in the past decade. There is no existing work on selection of potential nonlinear effects for subdistribution hazard model. We propose a two-stage procedure to select the linear and nonlinear covariate(s) simultaneously and estimate the selected covariate effect(s). We use spectral decomposition approach to distinguish the linear and nonlinear parts of each covariate and adaptive LASSO to select each of the 2 components. Extensive numerical studies are conducted to demonstrate that the proposed procedure can achieve good selection accuracy in the first stage and small estimation biases in the second stage. The proposed method is applied to analyze a cardiovascular disease data set with competing death causes.

AB - Subdistribution hazard model for competing risks data has been applied extensively in clinical researches. Variable selection methods of linear effects for competing risks data have been studied in the past decade. There is no existing work on selection of potential nonlinear effects for subdistribution hazard model. We propose a two-stage procedure to select the linear and nonlinear covariate(s) simultaneously and estimate the selected covariate effect(s). We use spectral decomposition approach to distinguish the linear and nonlinear parts of each covariate and adaptive LASSO to select each of the 2 components. Extensive numerical studies are conducted to demonstrate that the proposed procedure can achieve good selection accuracy in the first stage and small estimation biases in the second stage. The proposed method is applied to analyze a cardiovascular disease data set with competing death causes.

KW - Cubic b-spline

KW - Penalized log-likelihood

KW - Spectral decomposition

KW - Subdistribution hazard

UR - http://www.scopus.com/inward/record.url?scp=85044431633&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85044431633&partnerID=8YFLogxK

U2 - 10.1002/sim.7637

DO - 10.1002/sim.7637

M3 - Article

AN - SCOPUS:85044431633

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

ER -