Abstract :
Estimators of the reliability function in a GLM (generalized life model) are considered. The class of the GLM includes (among others) the Weibull, Pareto, Beta, Gompertz, and Rayleigh distribution. A proper general prior density and the predictive function for general class of distribution proposed by Al-Hussaini(1999) are used to obtain the exact estimate. Also, the Bayes estimates relative to symmetric loss function (quadratic loss), and asymmetric loss function (LINEX loss, and GE loss), are obtained. Comparisons are made between those estimators and the MLE applying to the Burr-XII model using the Bayes approximation due to Lindley. Monte Carlo simulation was used
Keywords :
Bayes methods; Monte Carlo methods; life testing; parameter estimation; reliability; Bayes estimates; Beta distribution; Burr-XII model; GE loss; Gompertz distribution; LINEX loss; Lindley approximation; Monte Carlo simulation; Pareto distribution; Rayleigh distribution; Weibull distribution; asymmetric loss function; completely censored samples; generalized life model; predictive function; prior density; quadratic loss; reliability function estimation; symmetric loss function; type-2 censored samples; Distribution functions; Entropy; Life estimation; Maximum likelihood estimation; Performance loss; Predictive models; Probability density function; Reliability theory; Shape; Testing;
