Information loss in a competing risks model

Research output: Contribution to journalArticle

Abstract

A semi-parametric model is investigated for the analysis of competing risks with covariates in the presence of independent random censoring. The information of the maximum partial likelihood estimates is compared with that of the maximum likelihood estimates, assuming the baseline hazard function is an unknown constant. The difference between the two information matrices is taken as the amount of information lost as the result of using the partial likelihood. It is a positive semi-definite matrix. A condition is given for no loss of information.

Original languageEnglish (US)
Pages (from-to)3211-3226
Number of pages16
JournalCommunications in Statistics - Theory and Methods
Volume23
Issue number11
DOIs
StatePublished - Jan 1 1994

Fingerprint

Competing Risks Model
Information Loss
Partial Likelihood
Random Censoring
Competing Risks
Positive Semidefinite Matrix
Information Matrix
Hazard Function
Semiparametric Model
Maximum Likelihood Estimate
Maximum Likelihood
Covariates
Baseline
Unknown
Estimate

All Science Journal Classification (ASJC) codes

  • Statistics and Probability

Cite this

Information loss in a competing risks model. / Wan, Jim.

In: Communications in Statistics - Theory and Methods, Vol. 23, No. 11, 01.01.1994, p. 3211-3226.

Research output: Contribution to journalArticle

@article{7fcc851817ab4a8cb277e8b28c1f3366,
title = "Information loss in a competing risks model",
abstract = "A semi-parametric model is investigated for the analysis of competing risks with covariates in the presence of independent random censoring. The information of the maximum partial likelihood estimates is compared with that of the maximum likelihood estimates, assuming the baseline hazard function is an unknown constant. The difference between the two information matrices is taken as the amount of information lost as the result of using the partial likelihood. It is a positive semi-definite matrix. A condition is given for no loss of information.",
author = "Jim Wan",
year = "1994",
month = "1",
day = "1",
doi = "10.1080/03610929408831442",
language = "English (US)",
volume = "23",
pages = "3211--3226",
journal = "Communications in Statistics - Theory and Methods",
issn = "0361-0926",
publisher = "Taylor and Francis Ltd.",
number = "11",

}

TY - JOUR

T1 - Information loss in a competing risks model

AU - Wan, Jim

PY - 1994/1/1

Y1 - 1994/1/1

N2 - A semi-parametric model is investigated for the analysis of competing risks with covariates in the presence of independent random censoring. The information of the maximum partial likelihood estimates is compared with that of the maximum likelihood estimates, assuming the baseline hazard function is an unknown constant. The difference between the two information matrices is taken as the amount of information lost as the result of using the partial likelihood. It is a positive semi-definite matrix. A condition is given for no loss of information.

AB - A semi-parametric model is investigated for the analysis of competing risks with covariates in the presence of independent random censoring. The information of the maximum partial likelihood estimates is compared with that of the maximum likelihood estimates, assuming the baseline hazard function is an unknown constant. The difference between the two information matrices is taken as the amount of information lost as the result of using the partial likelihood. It is a positive semi-definite matrix. A condition is given for no loss of information.

UR - http://www.scopus.com/inward/record.url?scp=84963165263&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84963165263&partnerID=8YFLogxK

U2 - 10.1080/03610929408831442

DO - 10.1080/03610929408831442

M3 - Article

VL - 23

SP - 3211

EP - 3226

JO - Communications in Statistics - Theory and Methods

JF - Communications in Statistics - Theory and Methods

SN - 0361-0926

IS - 11

ER -