Numerical Methods on the Triple Informative Prior Distribution for the Failure Rate Basic Gompertz Model

Abstract

In this paper, it has been dealt with basic Gompertz distribution. The maximum likelihood, Bayes methods of estimation were used to estimate the unknown shape parameter. The failure rate (hazard) function with the least loss was found using different priors (Gamma, exponential, chi-square and triple prior) under symmetric loss function (Degroot loss function). A comparison was made about the performance of these estimators with the numerical solution that was found using expansion methods (Bernstein polynomial and power function) which was applied to find the failure rate function numerically. The proficiency test of the proposed methods was conducted with a number of test examples. Finally, for computations the Matlab (R2015b) is used.