Biproportional scaling of matrices and the iterative proportional fitting procedure

A short proof is given of the necessary and sufficient conditions for the convergence of the Iterative Proportional Fitting procedure. The input consists of a nonnegative matrix and of positive target marginals for row sums and for column sums. The output is a sequence of scaled matrices to approximate the biproportional fit, that is, the scaling of the input matrix by means of row and column divisors in order to fit row and column sums to target marginals. Generally it is shown that certain structural properties of a biproportional scaling do not depend on the particular sequence used to approximate it. Specifically, the sequence that emerges from the Iterative Proportional Fitting procedure is analyzed by means of the L ₁-error that measures how current row and column sums compare to their target marginals. As a new result a formula for the limiting L ₁-error is obtained. The formula is in terms of partial sums of the target marginals, and easily yields the other well-known convergence characterizations.     

A structure theorem on bivariate positive quadrant dependent distributions and tests for independence in two-way contingency tables

In this paper, the set of all bivariate positive quadrant dependent distributions with fixed marginals is shown to be compact and convex. Extreme points of this convex set are enumerated in some specific examples. Applications are given in testing the hypothesis of independence against strict positive quadrant dependence in the context of ordinal contingency tables. The performance of two tests, one of which is based on eigenvalues of a random matrix, is compared. Various procedures based upon certain functions of the eigenvalues of a random matrix are also proposed for testing for independence in a two-way contingency table when the marginals are random.     

A type of Strassen's theorem for positive vector measures with values in dual spaces

In this paper, we extend a type of Strassen's theorem for the existence of probability measures with given marginals to positive vector measures with values in the dual of a barreled locally convex space which has certain order conditions. In this process of the extension we also give some useful properties for vector measures with values in dual spaces.

     

Conditional Iterative Proportional Fitting for Gaussian Distributions

A Gaussian version of the iterative proportional fitting procedure (IFP-P) was applied by Speed and Kiiveri to solve the likelihood equations in graphical Gaussian models. The calculation of the maximum likelihood estimates can be seen as the problem to find a Gaussian distribution with prescribed Gaussian marginals. We extend the Gaussian version of the IPF-P so that additionally given conditionals of Gaussian type are taken into account. The convergence of both proposed procedures, called conditional iterative proportional fitting procedures (CIPF-P), is proved.     

Dependence between two multivariate extremes

We extend the characterizations given by Takahashi (1988) for the independence and the total dependence of the univariate marginals of a multivariate extreme value distribution to its multivariate marginals. We also deal with the problem of how to measure the strength of the dependence among multivariate extremes. By presenting new definitions for the extremal coefficient, we propose measures that summarize the dependence between two multivariate extreme value distributions and preserve the main properties of the known bivariate coefficient for two univariate extreme value distributions. Finally, we illustrate these contributions to model the dependence among multivariate marginals with examples.     

The inverted complex Wishart distribution and its application to spectral estimation

The inverted complex Wishart distribution and its use for the construction of spectral estimates are studied. The density, some marginals of the distribution, and the first- and second-order moments are given. For a vector-valued time series, estimation of the spectral density at a collection of frequencies and estimation of the increments of the spectral distribution function in each of a set of frequency bands are considered. A formal procedure applies Bayes theorem, where the complex Wishart is used to represent the distribution of an average of adjacent periodogram values. A conjugate prior distribution for each parameter is an inverted complex Wishart distribution. Use of the procedure for estimation of a 2 × 2 spectral density matrix is discussed.     

关于内射模的P-平坦性

通过引用P-平坦模的定义,引入了右IPF环的概念,推广了右IF环的概念,这对研究IF环及QF环具有重要的作用,同时对右IPF环的性质作了一些刻画,得到了右IPF环的若干个等价命题;最后,用P-平坦模及右IPF环推出了正则环的一些等价条件．     

Regression models for multivariate ordered responses via the Plackett distribution

We investigate the properties of a class of discrete multivariate distributions whose univariate marginals have ordered categories, all the bivariate marginals, like in the Plackett distribution, have log-odds ratios which do not depend on cut points and all higher-order interactions are constrained to 0. We show that this class of distributions may be interpreted as a discretized version of a multivariate continuous distribution having univariate logistic marginals. Convenient features of this class relative to the class of ordered probit models (the discretized version of the multivariate normal) are highlighted. Relevant properties of this distribution like quadratic log-linear expansion, invariance to collapsing of adjacent categories, properties related to positive dependence, marginalization and conditioning are discussed briefly. When continuous explanatory variables are available, regression models may be fitted to relate the univariate logits (as in a proportional odds model) and the log-odds ratios to covariates.     

Extremal measures with prescribed marginals (finite case)

The problem of describing the extremal mesures with given marginals on a finite Cartesian product is studied.     

On the Distribution of a Sum of Sarmanov Distributed Random Variables

Built from given marginals with a flexible dependency structure, Sarmanov’s family of multivariate distributions became of interest in various fields. In this paper, we present some formulas for the density of the sum of several random variables joined by Sarmanov’s distribution, with accent on the particular case of exponentially distributed marginals. Such results are useful in solving, e.g., financial and actuarial problems.     

Three general multivariate semi-Pareto distributions and their characterizations

Three general multivariate semi-Pareto distributions are developed in this paper. First one—GMP^(k)(III) has univariate Pareto (III) marginals, it is characterized by the minimum of two independent and identically distributed random vectors. Second one—GMSP has univariate semi-Pareto marginals and it is characterized by finite sample minima. Third one—MSP is characterized through a geometric minimization procedure. All these three characterizations are based on the general and the particular solutions of the Euler's functional equations of k-variates.     

On multivariate infinitely divisible distributions

Simple conditions are given which characterize the generating function of a nonnegative multivariate infinitely divisible random vector. Necessary conditions on marginals, linear combinations, tail behavior, and zeroes are discussed, and a sufficient condition is given. The latter condition, which is a multivariate generalization of ordinary log-convexity, is shown to characterize only certain products of univariate infinitely divisible distributions.     

A class of two-type point processes

Summary A construction is given of a class of two-type point processes with Poisson marginals but possibly negative correlation between points of different types. Examples of the construction are given. The correlation structure of the processes is determined, and criteria obtained for the processes to be stationary, ergodic and mixing.Research supported by Social Science Research Council Spatial Data Project     

Construction of multivariate distributions with given marginals

Summary We make some remarks on the problem how to construct probability measures with given marginals. Questions of this kind arise if one wants to build a stochastic model in a situation where one has some idea of the kind of dependence and knows exactly certain marginal distributions.     

Bounds for functions of multivariate risks

Li et al. [Distributions with Fixed Marginals and Related Topics, vol. 28, Institute of Mathematics and Statistics, Hayward, CA, 1996, pp. 198-212] provide bounds on the distribution and on the tail for functions of dependent random vectors having fixed multivariate marginals. In this paper, we correct a result stated in the above article and we give improved bounds in the case of the sum of identically distributed random vectors. Moreover, we provide the dependence structures meeting the bounds when the fixed marginals are uniformly distributed on the k-dimensional hypercube. Finally, a definition of a multivariate risk measure is given along with actuarial/financial applications.     

Joint characteristic functions construction via copulas

When modelling dependent risks it is important to be able to generate joint distributions with given marginals. One of the ways which may be useful in connection with using the Fast Fourier Transform is to construct joint characteristic functions from marginal characteristic functions. In this paper a class of n-dimensional continuous copulas is presented which in turn lead to a simple construction of joint characteristic functions with given marginal characteristic functions. Bounds on various measures of correlation are also given.     

A class of bivariate distributions including the bivariate logistic

A univariate logistic distribution can be specified by considering a suitable form for the odds in favor of a failure against survival. This concept is extended to the bivariate case and a class of distributions, indexed by a parameter of association, having given marginals is proposed. Some properties of the proposed class of distributions are studied.     

Evaluation of nondominated solution sets for k-objective optimization problems: An exact method and approximations

Integrated Preference Functional (IPF) is a set functional that, given a discrete set of points for a multiple objective optimization problem, assigns a numerical value to that point set. This value provides a quantitative measure for comparing different sets of points generated by solution procedures for difficult multiple objective optimization problems. We introduced the IPF for bi-criteria optimization problems in [Carlyle, W.M., Fowler, J.W., Gel, E., Kim, B., 2003. Quantitative comparison of approximate solution sets for bi-criteria optimization problems. Decision Sciences 34 (1), 63–82]. As indicated in that paper, the computational effort to obtain IPF is negligible for bi-criteria problems. For three or more objective function cases, however, the exact calculation of IPF is computationally demanding, since this requires k (⩾3) dimensional integration.In this paper, we suggest a theoretical framework for obtaining IPF for k (⩾3) objectives. The exact method includes solving two main sub-problems: (1) finding the optimality region of weights for all potentially optimal points, and (2) computing volumes of k dimensional convex polytopes. Several different algorithms for both sub-problems can be found in the literature. We use existing methods from computational geometry (i.e., triangulation and convex hull algorithms) to develop a reasonable exact method for obtaining IPF. We have also experimented with a Monte Carlo approximation method and compared the results to those with the exact IPF method.     

A note on the Sarmanov bivariate distributions

We investigate the maximum correlation for Sarmanov bivariate distributions with fixed marginals and strengthen the existing results in the literature. The improvement in the maximum correlation is significant. A characterization of the Sarmanov distribution via chi-square divergence is also given. This extends Nelsen [13] result about the Farlie-Gumbel-Morgenstern (FGM) distribution.     

A note on characterizations of multivariate stable distributions

Several characterizations of multivariate stable distributions together with a characterization of multivariate normal distributions and multivariate stable distributions with Cauchy marginals are given. These are related to some standard characterizations of marcinkiewicz.Research supported, in part, by the Air Force Office of Scientific Research under Contract AFOSR 84-0113. Reproduction in whole or part is permitted for any purpose of the United States Government.