Robust nonparametric estimation of monotone regression functions with interval-censored observations

Ying Zhang, Gang Cheng, Wanzhu Tu

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Nonparametric estimation of monotone regression functions is a classical problem of practical importance. Robust estimation of monotone regression functions in situations involving interval-censored data is a challenging yet unresolved problem. Herein, we propose a nonparametric estimation method based on the principle of isotonic regression. Using empirical process theory, we show that the proposed estimator is asymptotically consistent under a specific metric. We further conduct a simulation study to evaluate the performance of the estimator in finite sample situations. As an illustration, we use the proposed method to estimate the mean body weight functions in a group of adolescents after they reach pubertal growth spurt.

Original languageEnglish (US)
Pages (from-to)720-730
Number of pages11
JournalBiometrics
Volume72
Issue number3
DOIs
StatePublished - Sep 1 2016

Fingerprint

Monotone Regression
Censored Observations
Monotone Function
Robust Estimation
Nonparametric Estimation
Regression Function
Isotonic Regression
Estimator
Interval-censored Data
Interval
Empirical Process
Weight Function
Body Weight
Simulation Study
Metric
Evaluate
Growth
Estimate
body weight
methodology

Keywords

  • Consistency
  • Doubly censored data
  • Interval-censoring
  • Isotonic regression
  • Monotone regression function

ASJC Scopus subject areas

  • Statistics and Probability
  • Medicine(all)
  • Immunology and Microbiology(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Agricultural and Biological Sciences(all)
  • Applied Mathematics

Cite this

Robust nonparametric estimation of monotone regression functions with interval-censored observations. / Zhang, Ying; Cheng, Gang; Tu, Wanzhu.

In: Biometrics, Vol. 72, No. 3, 01.09.2016, p. 720-730.

Research output: Contribution to journalArticle

@article{0bd69c594804457faf9b56158082ae55,
title = "Robust nonparametric estimation of monotone regression functions with interval-censored observations",
abstract = "Nonparametric estimation of monotone regression functions is a classical problem of practical importance. Robust estimation of monotone regression functions in situations involving interval-censored data is a challenging yet unresolved problem. Herein, we propose a nonparametric estimation method based on the principle of isotonic regression. Using empirical process theory, we show that the proposed estimator is asymptotically consistent under a specific metric. We further conduct a simulation study to evaluate the performance of the estimator in finite sample situations. As an illustration, we use the proposed method to estimate the mean body weight functions in a group of adolescents after they reach pubertal growth spurt.",
keywords = "Consistency, Doubly censored data, Interval-censoring, Isotonic regression, Monotone regression function",
author = "Ying Zhang and Gang Cheng and Wanzhu Tu",
year = "2016",
month = "9",
day = "1",
doi = "10.1111/biom.12465",
language = "English (US)",
volume = "72",
pages = "720--730",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "3",

}

TY - JOUR

T1 - Robust nonparametric estimation of monotone regression functions with interval-censored observations

AU - Zhang, Ying

AU - Cheng, Gang

AU - Tu, Wanzhu

PY - 2016/9/1

Y1 - 2016/9/1

N2 - Nonparametric estimation of monotone regression functions is a classical problem of practical importance. Robust estimation of monotone regression functions in situations involving interval-censored data is a challenging yet unresolved problem. Herein, we propose a nonparametric estimation method based on the principle of isotonic regression. Using empirical process theory, we show that the proposed estimator is asymptotically consistent under a specific metric. We further conduct a simulation study to evaluate the performance of the estimator in finite sample situations. As an illustration, we use the proposed method to estimate the mean body weight functions in a group of adolescents after they reach pubertal growth spurt.

AB - Nonparametric estimation of monotone regression functions is a classical problem of practical importance. Robust estimation of monotone regression functions in situations involving interval-censored data is a challenging yet unresolved problem. Herein, we propose a nonparametric estimation method based on the principle of isotonic regression. Using empirical process theory, we show that the proposed estimator is asymptotically consistent under a specific metric. We further conduct a simulation study to evaluate the performance of the estimator in finite sample situations. As an illustration, we use the proposed method to estimate the mean body weight functions in a group of adolescents after they reach pubertal growth spurt.

KW - Consistency

KW - Doubly censored data

KW - Interval-censoring

KW - Isotonic regression

KW - Monotone regression function

UR - http://www.scopus.com/inward/record.url?scp=85028251109&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85028251109&partnerID=8YFLogxK

U2 - 10.1111/biom.12465

DO - 10.1111/biom.12465

M3 - Article

C2 - 26757488

AN - SCOPUS:85028251109

VL - 72

SP - 720

EP - 730

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 3

ER -