Authors: Bhupendra Singh K K Sharma Shubhi Rathi Gajraj Singh
Publish Date: 2011/02/06
Volume: 27, Issue: 1, Pages: 51-67
Abstract
In life testing experiments the skewed distributions like lognormal Weibull gamma and generalized gamma are the most suitable models for recording the failure time measurements In this paper a generalized version of lognormal distribution is proposed and its goodnessoffit for a randomly censored data set representing the remission times of bladder cancer patients has been demonstrated and compared with other lifetime models considered in the literature The PP plots of KaplanMeier estimator against the survival functions of the considered models are used to show the goodnessoffit A simulation study is also performed to estimate the parameters in both the classical and Bayesian setups
Keywords: