Variable selection for optimal treatment decision

Wenbin Lu, Hao Zhang, Donglin Zeng

Research output: Contribution to journalArticle

54 Citations (Scopus)

Abstract

In decision-making on optimal treatment strategies, it is of great importance to identify variables that are involved in the decision rule, i.e. those interacting with the treatment. Effective variable selection helps to improve the prediction accuracy and enhance the interpretability of the decision rule. We propose a new penalized regression framework which can simultaneously estimate the optimal treatment strategy and identify important variables. The advantages of the new approach include: (i) it does not require the estimation of the baseline mean function of the response, which greatly improves the robustness of the estimator; (ii) the convenient loss-based framework makes it easier to adopt shrinkage methods for variable selection, which greatly facilitates implementation and statistical inferences for the estimator. The new procedure can be easily implemented by existing state-of-art software packages like LARS. Theoretical properties of the new estimator are studied. Its empirical performance is evaluated using simulation studies and further illustrated with an application to an AIDS clinical trial.

Original languageEnglish (US)
Pages (from-to)493-504
Number of pages12
JournalStatistical Methods in Medical Research
Volume22
Issue number5
DOIs
StatePublished - Oct 2013
Externally publishedYes

Fingerprint

Variable Selection
Patient Selection
Decision Rules
Estimator
Penalized Regression
Decision Making
Acquired Immunodeficiency Syndrome
Software
Interpretability
Clinical Trials
Shrinkage
Statistical Inference
Software Package
Baseline
Simulation Study
Robustness
Prediction
Estimate
Strategy
Framework

Keywords

  • A-learning
  • optimal treatment strategy
  • personalized drugs
  • shrinkage method
  • variable selection

ASJC Scopus subject areas

  • Epidemiology
  • Health Information Management
  • Statistics and Probability

Cite this

Variable selection for optimal treatment decision. / Lu, Wenbin; Zhang, Hao; Zeng, Donglin.

In: Statistical Methods in Medical Research, Vol. 22, No. 5, 10.2013, p. 493-504.

Research output: Contribution to journalArticle

Lu, Wenbin ; Zhang, Hao ; Zeng, Donglin. / Variable selection for optimal treatment decision. In: Statistical Methods in Medical Research. 2013 ; Vol. 22, No. 5. pp. 493-504.
@article{4f6ea46922894a3591267c895ea5c7a8,
title = "Variable selection for optimal treatment decision",
abstract = "In decision-making on optimal treatment strategies, it is of great importance to identify variables that are involved in the decision rule, i.e. those interacting with the treatment. Effective variable selection helps to improve the prediction accuracy and enhance the interpretability of the decision rule. We propose a new penalized regression framework which can simultaneously estimate the optimal treatment strategy and identify important variables. The advantages of the new approach include: (i) it does not require the estimation of the baseline mean function of the response, which greatly improves the robustness of the estimator; (ii) the convenient loss-based framework makes it easier to adopt shrinkage methods for variable selection, which greatly facilitates implementation and statistical inferences for the estimator. The new procedure can be easily implemented by existing state-of-art software packages like LARS. Theoretical properties of the new estimator are studied. Its empirical performance is evaluated using simulation studies and further illustrated with an application to an AIDS clinical trial.",
keywords = "A-learning, optimal treatment strategy, personalized drugs, shrinkage method, variable selection",
author = "Wenbin Lu and Hao Zhang and Donglin Zeng",
year = "2013",
month = "10",
doi = "10.1177/0962280211428383",
language = "English (US)",
volume = "22",
pages = "493--504",
journal = "Statistical Methods in Medical Research",
issn = "0962-2802",
publisher = "SAGE Publications Ltd",
number = "5",

}

TY - JOUR

T1 - Variable selection for optimal treatment decision

AU - Lu, Wenbin

AU - Zhang, Hao

AU - Zeng, Donglin

PY - 2013/10

Y1 - 2013/10

N2 - In decision-making on optimal treatment strategies, it is of great importance to identify variables that are involved in the decision rule, i.e. those interacting with the treatment. Effective variable selection helps to improve the prediction accuracy and enhance the interpretability of the decision rule. We propose a new penalized regression framework which can simultaneously estimate the optimal treatment strategy and identify important variables. The advantages of the new approach include: (i) it does not require the estimation of the baseline mean function of the response, which greatly improves the robustness of the estimator; (ii) the convenient loss-based framework makes it easier to adopt shrinkage methods for variable selection, which greatly facilitates implementation and statistical inferences for the estimator. The new procedure can be easily implemented by existing state-of-art software packages like LARS. Theoretical properties of the new estimator are studied. Its empirical performance is evaluated using simulation studies and further illustrated with an application to an AIDS clinical trial.

AB - In decision-making on optimal treatment strategies, it is of great importance to identify variables that are involved in the decision rule, i.e. those interacting with the treatment. Effective variable selection helps to improve the prediction accuracy and enhance the interpretability of the decision rule. We propose a new penalized regression framework which can simultaneously estimate the optimal treatment strategy and identify important variables. The advantages of the new approach include: (i) it does not require the estimation of the baseline mean function of the response, which greatly improves the robustness of the estimator; (ii) the convenient loss-based framework makes it easier to adopt shrinkage methods for variable selection, which greatly facilitates implementation and statistical inferences for the estimator. The new procedure can be easily implemented by existing state-of-art software packages like LARS. Theoretical properties of the new estimator are studied. Its empirical performance is evaluated using simulation studies and further illustrated with an application to an AIDS clinical trial.

KW - A-learning

KW - optimal treatment strategy

KW - personalized drugs

KW - shrinkage method

KW - variable selection

UR - http://www.scopus.com/inward/record.url?scp=84886467569&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84886467569&partnerID=8YFLogxK

U2 - 10.1177/0962280211428383

DO - 10.1177/0962280211428383

M3 - Article

C2 - 22116341

AN - SCOPUS:84886467569

VL - 22

SP - 493

EP - 504

JO - Statistical Methods in Medical Research

JF - Statistical Methods in Medical Research

SN - 0962-2802

IS - 5

ER -