A two-step approach for variable selection in linear regression with measurement error

Jiyeon Song, Seung Jun Shin

Research output: Contribution to journalArticle

Abstract

It is important to identify informative variables in high dimensional data analysis; however, it becomes a challenging task when covariates are contaminated by measurement error due to the bias induced by measurement error. In this article, we present a two-step approach for variable selection in the presence of measurement error. In the first step, we directly select important variables from the contaminated covariates as if there is no measurement error. We then apply, in the following step, orthogonal regression to obtain the unbiased estimates of regression coefficients identified in the previous step. In addition, we propose a modification of the twostep approach to further enhance the variable selection performance. Various simulation studies demonstrate the promising performance of the proposed method.

Original languageEnglish
Pages (from-to)47-55
Number of pages9
JournalCommunications for Statistical Applications and Methods
Volume26
Issue number1
DOIs
Publication statusPublished - 2019 Jan 1

Fingerprint

Variable Selection
Measurement errors
Linear regression
Measurement Error
Covariates
Orthogonal Regression
High-dimensional Data
Regression Coefficient
Data analysis
Simulation Study
Variable selection
Measurement error
Estimate
Demonstrate

Keywords

  • Measurement error
  • Penalized orthogonal regression
  • SIMEX

ASJC Scopus subject areas

  • Statistics and Probability
  • Modelling and Simulation
  • Finance
  • Statistics, Probability and Uncertainty
  • Applied Mathematics

Cite this

A two-step approach for variable selection in linear regression with measurement error. / Song, Jiyeon; Shin, Seung Jun.

In: Communications for Statistical Applications and Methods, Vol. 26, No. 1, 01.01.2019, p. 47-55.

Research output: Contribution to journalArticle

@article{f7b1a535683942bf9a00d3863caec0a6,
title = "A two-step approach for variable selection in linear regression with measurement error",
abstract = "It is important to identify informative variables in high dimensional data analysis; however, it becomes a challenging task when covariates are contaminated by measurement error due to the bias induced by measurement error. In this article, we present a two-step approach for variable selection in the presence of measurement error. In the first step, we directly select important variables from the contaminated covariates as if there is no measurement error. We then apply, in the following step, orthogonal regression to obtain the unbiased estimates of regression coefficients identified in the previous step. In addition, we propose a modification of the twostep approach to further enhance the variable selection performance. Various simulation studies demonstrate the promising performance of the proposed method.",
keywords = "Measurement error, Penalized orthogonal regression, SIMEX",
author = "Jiyeon Song and Shin, {Seung Jun}",
year = "2019",
month = "1",
day = "1",
doi = "10.29220/CSAM.2019.26.1.047",
language = "English",
volume = "26",
pages = "47--55",
journal = "Communications for Statistical Applications and Methods",
issn = "2287-7843",
publisher = "Korean Statistical Society",
number = "1",

}

TY - JOUR

T1 - A two-step approach for variable selection in linear regression with measurement error

AU - Song, Jiyeon

AU - Shin, Seung Jun

PY - 2019/1/1

Y1 - 2019/1/1

N2 - It is important to identify informative variables in high dimensional data analysis; however, it becomes a challenging task when covariates are contaminated by measurement error due to the bias induced by measurement error. In this article, we present a two-step approach for variable selection in the presence of measurement error. In the first step, we directly select important variables from the contaminated covariates as if there is no measurement error. We then apply, in the following step, orthogonal regression to obtain the unbiased estimates of regression coefficients identified in the previous step. In addition, we propose a modification of the twostep approach to further enhance the variable selection performance. Various simulation studies demonstrate the promising performance of the proposed method.

AB - It is important to identify informative variables in high dimensional data analysis; however, it becomes a challenging task when covariates are contaminated by measurement error due to the bias induced by measurement error. In this article, we present a two-step approach for variable selection in the presence of measurement error. In the first step, we directly select important variables from the contaminated covariates as if there is no measurement error. We then apply, in the following step, orthogonal regression to obtain the unbiased estimates of regression coefficients identified in the previous step. In addition, we propose a modification of the twostep approach to further enhance the variable selection performance. Various simulation studies demonstrate the promising performance of the proposed method.

KW - Measurement error

KW - Penalized orthogonal regression

KW - SIMEX

UR - http://www.scopus.com/inward/record.url?scp=85065240172&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85065240172&partnerID=8YFLogxK

U2 - 10.29220/CSAM.2019.26.1.047

DO - 10.29220/CSAM.2019.26.1.047

M3 - Article

AN - SCOPUS:85065240172

VL - 26

SP - 47

EP - 55

JO - Communications for Statistical Applications and Methods

JF - Communications for Statistical Applications and Methods

SN - 2287-7843

IS - 1

ER -