Academic Journals Database
Disseminating quality controlled scientific knowledge

Bayesian and Non-Bayesian Estimation of the Inverse Weibull Model Based on Generalized Order Statistics

ADD TO MY LIST
 
Author(s): Ahmed H. Abd Ellah

Journal: Intelligent Information Management
ISSN 2160-5912

Volume: 04;
Issue: 02;
Start page: 23;
Date: 2012;
Original page

Keywords: Inverse Weibull Distribution | Generalized Order Statistics | Record Values | Progressive Type-II Censored | Balanced Type Loss Function | Bootstrap Estimation

ABSTRACT
The concept of generalized order statistics has been introduced as a unified approach to a variety of models of ordered random variables with different interpretations. In this paper, we develop methodology for constructing inference based on n selected generalized order statistics (GOS) from inverse Weibull distribution (IWD), Bayesian and non-Bayesian approaches have been used to obtain the estimators of the parameters and reliability function. We have examined Bayes estimates under various losses such as the balanced squared error (balanced SEL) and balanced LINEX loss functions are considered. We show that Bayes estimate under balanced SEL and balanced LINEX loss functions are more general, which include the symmetric and asymmetric losses as special cases. This was done under assumption of discrete-continuous mixture prior for the unknown model parameters. The parametric bootstrap method has been used to construct confidence interval for the parameters and reliability function. Progressively type-II censored and k-record values as a special case of GOS are considered. Finally a practical example using real data set was used for illustration.
Why do you need a reservation system?      Save time & money - Smart Internet Solutions