Approximate Bayesian Credible Intervals for the Parameters of the Gompertz Distribution with Failure Censored Data
摘要
We considered Bayesian interval estimation for the parameters of the Gompertz distribution with failure-censored data. We derived credible intervals for the shape and scale parameters using the Metropolis–Hastings algorithm and higher-order asymptotic approximations of the marginal posterior distribution under several choices of prior distributions. We assessed the performance of the credible intervals, comparing them to likelihood-based intervals with a focus on error probabilities. To illustrate the findings, we present a real-world data example.