ON COMPARISON OF ESTIMATORS IN A GENERALIZED LIFE MODEL

Bayes estimation of the scale parameter in a generalized life distribution is considered under entropy loss, linex loss and squared error loss (SEL) respectively. Risks and Bayes risks for those estimators as well as for maximum likelihood estimator (MLE), median unbiased estimator (MUE) are calculated and simultaneously compared for a number of priors. Optimal estimator relative to a particular loss function and prior are selected individually. Optimal estimators in Zacks and Even's sense for the scale parameter and reliability function are also obtained separately. Numerical studies indicate that the optimal estimators depend on the comparison criteria.