The characterization of distribution is useful for selection of adequate distribution to describe the observed values obtained in an experiment and is one of the methods of finding the distribution. Chapter 3 and 4 are concerned with the characterization developed by Kemp and Kemp (2004) and Ahmad and Roohi (2004). In Chapter 5, the recurrence relations between ordinary moments are established. A general characterization theorem, based on recurrence relation of ordinary moments is derived for a general class of discrete distributions. Chapter 6 deals with the recursive relations of factorial moments obtained by successive differentiation of factorial moment generating functions. In Chapters 7, 8, and 9 the theorems are then applied to numerous discrete probability distributions to provide specific characterizations for each one of them. Since information concerning moments is more often available than the knowledge of probability distribution as a whole, we expect these properties to be useful in dealing with the practical problems.

If X and Y are random variables such that P (X> Y) = 1 and the conditional distribution of Y given X is binomial, then Moran (1952) showed that Y and (X-Y) are independent if X is Poisson. This document extends Moran's result to a more general type of conditional distribution of Y given X, using only partial independence of Y and X-Y. This provides a generalization of a recent results of Janardhan and Rao (1982) on the characterization of generalized Polya-Eggenberger distribution. A variant of Moran's theorem is proved which generalizes the results of Patil and Seshadri (1964) on the characterization of the distribution of a random variable x based on some conditions on the conditional distribution of Y given X and the independence of Y and X-Y.

Provides in an organized manner characterizations of univariate probability distributions with many new results published in this area since the 1978 work of Golambos & Kotz "Characterizations of Probability Distributions" (Springer), together with applications of the theory in model fitting and predictions.

Characterization theorems in probability theory and mathematical statistics are such theorems that establish a connection between the type of the distribution of random variables or random vectors and certain general properties of functions in them. For example, the assumption that two linear (or non-linear) statistics are identically distributed (or independent, or have a constancy regression and so on) can be used to characterize various populations. Verification of conditions of this or that characterization theorem in practice is possible only with some error, i.e., only to a certain degree of accuracy. Such a situation is observed, for instance, in the cases where a sample of finite size is considered. That is why there arises the following natural question. Suppose that the conditions of the characterization theorem are fulfilled not exactly but only approximately. May we assert that the conclusion of the theorem is also fulfilled approximately? Questions of this kind give rise to a following problem: determine the degree of realizability of the conclusions of mathematical statements in the case of approximate validity of conditions.