Conditional lifetime data analysis using the limited expected value function

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Much failure, and other event, data are commonly highly censored. Consequently this limits the efficacy of many statistical analysis techniques. The limited expected value (LEV) function presents an alternative way of characterizing lifetime distributions. In essence the LEV provides a means of calculating a truncated mean time to failure (MTTF) (or mean time before failure (MTBF) if appropriate) that is adjusted at each of the censoring times and so appears potentially suitable for dealing with censored data structures. In theory, the LEV has been defined for many standard distributions, however its practical use is not well developed. This paper aims to extend the theory of LEV for typical censoring structures to develop procedures that will assist in model identification as well as parameter estimation. Applications to typical event data will be presented and the use of LEV in comparison with a selection of existing lifetime distributional analysis will be made based on some preliminary research.
LanguageEnglish
Pages185-192
Number of pages7
JournalQuality and Reliability Engineering International
Volume20
Issue number3
DOIs
Publication statusPublished - 2004

Fingerprint

Parameter estimation
Data structures
Statistical methods
Identification (control systems)
Expected value
Value function
Censoring

Keywords

  • data analysis
  • censorship
  • reliability engineering
  • statistics
  • quality management

Cite this

@article{1e6f2e510060481b8f30b2480d04d282,
title = "Conditional lifetime data analysis using the limited expected value function",
abstract = "Much failure, and other event, data are commonly highly censored. Consequently this limits the efficacy of many statistical analysis techniques. The limited expected value (LEV) function presents an alternative way of characterizing lifetime distributions. In essence the LEV provides a means of calculating a truncated mean time to failure (MTTF) (or mean time before failure (MTBF) if appropriate) that is adjusted at each of the censoring times and so appears potentially suitable for dealing with censored data structures. In theory, the LEV has been defined for many standard distributions, however its practical use is not well developed. This paper aims to extend the theory of LEV for typical censoring structures to develop procedures that will assist in model identification as well as parameter estimation. Applications to typical event data will be presented and the use of LEV in comparison with a selection of existing lifetime distributional analysis will be made based on some preliminary research.",
keywords = "data analysis, censorship, reliability engineering, statistics, quality management",
author = "J.L. Quigley and L.A. Walls",
year = "2004",
doi = "10.1002/qre.559",
language = "English",
volume = "20",
pages = "185--192",
journal = "Quality and Reliability Engineering International",
issn = "0748-8017",
number = "3",

}

TY - JOUR

T1 - Conditional lifetime data analysis using the limited expected value function

AU - Quigley, J.L.

AU - Walls, L.A.

PY - 2004

Y1 - 2004

N2 - Much failure, and other event, data are commonly highly censored. Consequently this limits the efficacy of many statistical analysis techniques. The limited expected value (LEV) function presents an alternative way of characterizing lifetime distributions. In essence the LEV provides a means of calculating a truncated mean time to failure (MTTF) (or mean time before failure (MTBF) if appropriate) that is adjusted at each of the censoring times and so appears potentially suitable for dealing with censored data structures. In theory, the LEV has been defined for many standard distributions, however its practical use is not well developed. This paper aims to extend the theory of LEV for typical censoring structures to develop procedures that will assist in model identification as well as parameter estimation. Applications to typical event data will be presented and the use of LEV in comparison with a selection of existing lifetime distributional analysis will be made based on some preliminary research.

AB - Much failure, and other event, data are commonly highly censored. Consequently this limits the efficacy of many statistical analysis techniques. The limited expected value (LEV) function presents an alternative way of characterizing lifetime distributions. In essence the LEV provides a means of calculating a truncated mean time to failure (MTTF) (or mean time before failure (MTBF) if appropriate) that is adjusted at each of the censoring times and so appears potentially suitable for dealing with censored data structures. In theory, the LEV has been defined for many standard distributions, however its practical use is not well developed. This paper aims to extend the theory of LEV for typical censoring structures to develop procedures that will assist in model identification as well as parameter estimation. Applications to typical event data will be presented and the use of LEV in comparison with a selection of existing lifetime distributional analysis will be made based on some preliminary research.

KW - data analysis

KW - censorship

KW - reliability engineering

KW - statistics

KW - quality management

UR - http://dx.doi.org/10.1002/qre.559

U2 - 10.1002/qre.559

DO - 10.1002/qre.559

M3 - Article

VL - 20

SP - 185

EP - 192

JO - Quality and Reliability Engineering International

T2 - Quality and Reliability Engineering International

JF - Quality and Reliability Engineering International

SN - 0748-8017

IS - 3

ER -