Atlantis Press Paper Details: A Study on Imputing Censored Observations for Exponential Distribution Based on Random Censoring

Censoring models are frequently employed in reliability analysis to reduce experimental time. There are three censoring model: type-I, type-II and random censoring. In this study, we focus on the right-random censoring model. In the previous literature, an imputation of the censored observation is considered as the censoring time (Miller (1981), Lawless (1982), Lee (1992) and among others). Clearly, the censored observation is imputed by the censoring time to underestimate the original failure time. In this paper, we consider the failure time to follow an exponential distribution, and alternatively we attempt to propose three measures to impute the censored observations. By Monte Carlo simulation, the goodness of fit test is employed to compare the four methods of imputing censored observations. It is found that the method of imputing censored data by censoring time little outperforms the other three imputing methods.

		back to session results

title:		A Study on Imputing Censored Observations for Exponential Distribution Based on Random Censoring
publication:		JCIS-2006 Proceedings
part of series:		Advances in Intelligent Systems Research
ISBN:		978-90-78677-01-7
ISSN:		1951-6851
DOI:		doi:10.2991/jcis.2006.119 (how to use a DOI)
author(s):		Kuo-Ching Chiou
corresponding author:		Kuo-Ching Chiou
publication date:		October 2006
keywords:		Censoring time, Imputation value, exponential distribution
abstract:		Censoring models are frequently employed in reliability analysis to reduce experimental time. There are three censoring model: type-I, type-II and random censoring. In this study, we focus on the right-random censoring model. In the previous literature, an imputation of the censored observation is considered as the censoring time (Miller (1981), Lawless (1982), Lee (1992) and among others). Clearly, the censored observation is imputed by the censoring time to underestimate the original failure time. In this paper, we consider the failure time to follow an exponential distribution, and alternatively we attempt to propose three measures to impute the censored observations. By Monte Carlo simulation, the goodness of fit test is employed to compare the four methods of imputing censored observations. It is found that the method of imputing censored data by censoring time little outperforms the other three imputing methods.
copyright:		© Atlantis Press. This article is distributed under the terms of the Creative Commons Attribution License, which permits non-commercial use, distribution and reproduction in any medium, provided the original work is properly cited.
full text:		JCIS06-CSI-44.pdf (462 K)