Semiparametric maximum likelihood inference for truncated or biased-sampling data

Hao Liu, Jing Ning, Jing Qin, Yu Shen

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

Sample selection bias has long been recognized in many fields including clinical trials, epidemiology studies, genome-wide association studies, and wildlife management. This paper investigates the maximum likelihood estimation for censored survival data with selection bias under the Cox regression models where the selection process is modeled parametrically. A novel expectation-maximization algorithm is proposed and shown to have considerable computational advantages. Rigorous asymptotic properties of the estimator are established. Extensive simulation studies and a data analysis are conducted to investigate the performance of the proposed estimation procedure.

Original languageEnglish (US)
Pages (from-to)1087-1115
Number of pages29
JournalStatistica Sinica
Volume26
Issue number3
DOIs
StatePublished - Jul 1 2016
Externally publishedYes

Fingerprint

Biased Sampling
Selection Bias
Likelihood Inference
Maximum Likelihood
Cox Regression Model
Censored Survival Data
Sample Selection
Expectation-maximization Algorithm
Epidemiology
Maximum Likelihood Estimation
Clinical Trials
Asymptotic Properties
Data analysis
Genome
Simulation Study
Estimator
Wildlife management
Maximum likelihood
Sampling
Regression model

Keywords

  • Biased sampling
  • Length bias data
  • Truncated and rightcensored survival data

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Semiparametric maximum likelihood inference for truncated or biased-sampling data. / Liu, Hao; Ning, Jing; Qin, Jing; Shen, Yu.

In: Statistica Sinica, Vol. 26, No. 3, 01.07.2016, p. 1087-1115.

Research output: Contribution to journalArticle

Liu, Hao ; Ning, Jing ; Qin, Jing ; Shen, Yu. / Semiparametric maximum likelihood inference for truncated or biased-sampling data. In: Statistica Sinica. 2016 ; Vol. 26, No. 3. pp. 1087-1115.
@article{c951a7cf165248d78927ee1b9c074af7,
title = "Semiparametric maximum likelihood inference for truncated or biased-sampling data",
abstract = "Sample selection bias has long been recognized in many fields including clinical trials, epidemiology studies, genome-wide association studies, and wildlife management. This paper investigates the maximum likelihood estimation for censored survival data with selection bias under the Cox regression models where the selection process is modeled parametrically. A novel expectation-maximization algorithm is proposed and shown to have considerable computational advantages. Rigorous asymptotic properties of the estimator are established. Extensive simulation studies and a data analysis are conducted to investigate the performance of the proposed estimation procedure.",
keywords = "Biased sampling, Length bias data, Truncated and rightcensored survival data",
author = "Hao Liu and Jing Ning and Jing Qin and Yu Shen",
year = "2016",
month = "7",
day = "1",
doi = "10.5705/ss.2014.094",
language = "English (US)",
volume = "26",
pages = "1087--1115",
journal = "Statistica Sinica",
issn = "1017-0405",
publisher = "Institute of Statistical Science",
number = "3",

}

TY - JOUR

T1 - Semiparametric maximum likelihood inference for truncated or biased-sampling data

AU - Liu, Hao

AU - Ning, Jing

AU - Qin, Jing

AU - Shen, Yu

PY - 2016/7/1

Y1 - 2016/7/1

N2 - Sample selection bias has long been recognized in many fields including clinical trials, epidemiology studies, genome-wide association studies, and wildlife management. This paper investigates the maximum likelihood estimation for censored survival data with selection bias under the Cox regression models where the selection process is modeled parametrically. A novel expectation-maximization algorithm is proposed and shown to have considerable computational advantages. Rigorous asymptotic properties of the estimator are established. Extensive simulation studies and a data analysis are conducted to investigate the performance of the proposed estimation procedure.

AB - Sample selection bias has long been recognized in many fields including clinical trials, epidemiology studies, genome-wide association studies, and wildlife management. This paper investigates the maximum likelihood estimation for censored survival data with selection bias under the Cox regression models where the selection process is modeled parametrically. A novel expectation-maximization algorithm is proposed and shown to have considerable computational advantages. Rigorous asymptotic properties of the estimator are established. Extensive simulation studies and a data analysis are conducted to investigate the performance of the proposed estimation procedure.

KW - Biased sampling

KW - Length bias data

KW - Truncated and rightcensored survival data

UR - http://www.scopus.com/inward/record.url?scp=85011911855&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85011911855&partnerID=8YFLogxK

U2 - 10.5705/ss.2014.094

DO - 10.5705/ss.2014.094

M3 - Article

AN - SCOPUS:85011911855

VL - 26

SP - 1087

EP - 1115

JO - Statistica Sinica

JF - Statistica Sinica

SN - 1017-0405

IS - 3

ER -