Maximum likelihood estimates of some probability model of discrete distributions

Abstract

In this work the new multivariate discrete probability model of distribution of random sums with unobserved components is proposed.The maximum likelihood estimates for this model are determined in the case that all the elements of the sample implementation, namely the observed sums of unobserved components have only singular partition. In the case, that some element of the sample implementation has more then one partition, it is not possible to establish the maximum likelihood estimates.