THE TRUNCATED XLINDLEY DISTRIBUTION WITH CLASSIC AND 6BAYESIAN INFERENCE UNDER CENSORED DATA

Abstract

We provide a brand-new distribution based on the model of Lindley, with an emphasis on the estimation of its unknown parameters. After introducing the new distribution, we cover two approaches to estimate its parameters; in the presence of a censored scheme, we first use a traditional approach, which is The maximum likelihood technique, then we use the Bayesian approach. The BarzilaiBrown algorithm is used to derive the censored maximum likelihood estimators while a Monte Carlo Markov chains (MCMC) procedure is applied to derive the Bayesian ones. Three loss functions are used to provide the Bayesian estimators: the entropy, the generalized quadratic, and the Linex functions. Using Pitman's proximity criteria; the maximum likelihood and the Bayesian estimations are compared. All of the provided estimations techniques have been evaluated throughout simulation studies. Finally, we consider two sample Bayes predictions to predict future order statistics