THE DOUBLE PRIOR SELECTION FOR THE PARAMETER OF INVERTED EXPONENTIAL DISTRIBUTION UNDER TYPE-II CENSORING

This paper is concerned with the comparison of double priors assumed for the parameter of inverted exponential distribution for quadratic loss function under Type-II censoring. The three different double priors viz. (i) inverted gamma and Gumble Type-II prior (ii) inverted gamma and Jeffrey's prior (iii) Gumble Type-II and Jeffrey's prior. The results obtained under the above double priors are also compared with that of based on only inverted gamma prior. Under Type-II censoring, Bayes estimation of the parameter and reliability at time t are obtained along with their equal tail credible intervals. The Bayes predictive estimator for the future observation and for the remaining order statistics after the censoring period are also derived along with their equal tail credible intervals. A simulation study is carried out to compare the performance of the Bayes estimators under the double prior distributions.