A NEW PROCEDURE : BAYESIAN SELECTION TO FIND THE BEST OF GEOMETRIC POPULATION UNDER GENERAL LOSS FUNCTION

Abstract

In many practical situations theexperimenter is confronted with the problem ofchoosing the best one of a number ofpopulations or categories or ranking themaccording to their performance . This paperderives a procedure for selecting the better ofTwo Geometric populations employing adecision-theoretic Bayesian framework withBeta prior under general loss function .the numerical results for this procedureare given by using Math Works Matlab ver 7.0.1with different loss functions constant , linear andquadratic , where in one equation we can obtainthe Bayes risk for the three types of the lossfunctions : constant , linear and quadratic .