The Empirical Bayes Estimator and Mixed Distributions
Nestor Ruben Barraza, Marcelo de Souza Lauretto, Carlos Alberto de Bragança Pereira, Julio Michael Stern · AIP conference proceedings · 2008
The empirical Bayes estimator of the probability of a successful event is deduced from mixed distributions. Specially, binomial and Poisson mixed distributions are analyzed. Some mixing distributions, well known in Reliability, Queuing Theory and other areas of Engineering, are considered. As it will be shown, a family of estimators with interesting characteristics is obtained for different mixing distributions. This family includes those well known estimators in the literature. As another result, estimators with non linear behavior as a function of the empirical sample are obtained.