A development of entropy optimization methods

Abstract

In this study by considering entropy optimization measure on the given set of moment vector functions special functional according to each measure is defined. Via the moment vector functions giving the least and the greatest values to the mentioned functionals, Generalized Entropy Optimization Distributions are obtained. Each of this distributions is closest (or remotest) to the given a priori distribution in the corresonding measure. Mentioned process of obtaining generalized optimi zation distributions and the establishment of required relations can be formulated as a development of entropy optimization methods. Moreover, some properties of entropy optimization operators, in the other words entropy optimization distributions considered on the compact set of moment vector functions are given.