Conditionally Reducible Natural Exponential Families and Enriched Conjugate Priors

Consider a standard conjugate family of prior distributions for a vector-parameter indexing an exponential family. Two distinct model parameterizations may well lead to standard conjugate families which are not consistent, i.e. one family cannot be derived from the other by the usual change-of-variable technique. This raises the problem of finding suitable parameterizations that may lead to enriched conjugate families which are more flexible than the traditional ones. The previous remark motivates the definition of a new property for an exponential family, named conditional reducibility. Features of conditionally-reducible natural exponential families are investigated thoroughly. In particular, we relate this new property to the notion of cut, and show that conditionally-reducible families admit a reparameterization in terms of a vector having likelihood-independent components. A general methodology to obtain enriched conjugate distributions for conditionally-reducible families is described in detail, generalizing previous works and more recent contributions in the area. The theory is illustrated with reference to natural exponential families having simple quadratic variance function.     

Minimum variance unbiased estimation in natural exponential families generated by stable distributions

The class of nature exponential families generated by stable distributions has been introduced in different contexts by several authors. Tweedie (1984) and Jorgensen (1987) studied this class in the context of generalized liner models and exponential dispersion models. Bar-Lev and Enis (1986) introduced this class in the context of the property of reproducibility in natural exponential families and Hougaard (1986) found the distributions in this class to be natural candidates for applications as survival distributions in life tables for heterogeneous populations. In this note, we consider such a class in the context of minimum variance unbiased estimation. For each family in this class, we obtain an explicit expression for the uniformly minimum variance unbiased estimator for the r-th cumlant, the density function, and the reliability function.     

On Unbiased Estimation of Positive Integral Powers of the Natural Parameter in Exponential Families

We explore the structure of one‐parameter exponential families admitting an unbiased estimator for a positive integral power of the natural parameter. It is seen that only exponential families dominated by Lebesgue measure can have this property. It is outlined that similar results can be obtained for other functions of the natural parameter.     

Instability, Sensitivity, and Degeneracy of Discrete Exponential Families

In applications to dependent data, first and foremost relational data, a number of discrete exponential family models has turned out to be near-degenerate and problematic in terms of Markov chain Monte Carlo simulation and statistical inference. We introduce the notion of instability with an eye to characterize, detect, and penalize discrete exponential family models that are near-degenerate and problematic in terms of Markov chain Monte Carlo simulation and statistical inference. We show that unstable discrete exponential family models are characterized by excessive sensitivity and near-degeneracy. In special cases, the subset of the natural parameter space corresponding to non-degenerate distributions and mean-value parameters far from the boundary of the mean-value parameter space turns out to be a lower-dimensional subspace of the natural parameter space. These characteristics of unstable discrete exponential family models tend to obstruct Markov chain Monte Carlo simulation and statistical inference. In applications to relational data, we show that discrete exponential family models with Markov dependence tend to be unstable and that the parameter space of some curved exponential families contains unstable subsets.     

Mixture of the Riesz distribution with respect to the generalized multivariate gamma distribution

Wishart natural exponential families (NEFs) characterized by Letac (1989) are extended to the Riesz NEFs on symmetric matrices. These families are characterized by their variance functions defined in Hassairi and Lajmi (2001). This work uses a particular basis of these NEFs to describe the class of the generalized multivariate gamma distributions and then to study the statistical model obtained by the mixture of this distribution with the Riesz one on the space of symmetric matrices.     

Stratified Gaussian graphical models

Gaussian graphical models represent the backbone of the statistical toolbox for analyzing continuous multivariate systems. However, due to the intrinsic properties of the multivariate normal distribution, use of this model family may hide certain forms of context-specific independence that are natural to consider from an applied perspective. Such independencies have been earlier introduced to generalize discrete graphical models and Bayesian networks into more flexible model families. Here, we adapt the idea of context-specific independence to Gaussian graphical models by introducing a stratification of the Euclidean space such that a conditional independence may hold in certain segments but be absent elsewhere. It is shown that the stratified models define a curved exponential family, which retains considerable tractability for parameter estimation and model selection.     

Bhattacharyya matrices and cubic exponential families

This paper introduces a notion of semi-diagonality for a Bhattacharyya matrix to give a characterization of the Letac–Mora class of real natural exponential families having cubic variance function.     

ON TESTING THE EQUALITY OF LOCATION PARAMETERS IN k CENSORED EXPONENTIAL DISTRIBUTIONS

Kumar and Patel (1971) have considered the problem of testing the equality of location parameters of two exponential distributions on the basis of samples censored from above, when the scale parameters are the same and unknown. The test proposed by them is shown to be biased for n₁≠n₂, while for n₁=n₂ the test possesses the property of monotonicity and is equivalent to the likelihood ratio test, which is considered by Epstein and Tsao (1953) and Dubey (1963a, 1963b). Epstein and Tsao state that the test is unbiased. We may note that when the scale parameters of k exponential distributions are unknown the problem of testing the equality of location parameters is reducible to that of testing the equality of parameters in k rectangular populations for which a test and its power function were given by Khatri (1960, 1965); Jaiswal (1969) considered similar problems in his thesis. Here we extend the problem of testing the equality of k exponential distributions on the basis of samples censored from above when the scale parameters are equal and unknown, and we establish the likelihood ratio test (LET) and the union-intersection test (UIT) procedures. Using the results previously derived by Jaiswal (1969), we obtain the power function for the LET and for k= 2 show that the test possesses the property of monotonicity. The power function of the UIT is also given.     

Bayesian inference on multiply sequential order statistics from heterogeneous exponential populations with GLR test for homogeneity

In this article, sequential order statistics (SOS) coming from heterogeneous exponential distributions are considered. Maximum likelihood and Bayesian estimates of parameters are derived on the basis of multiple SOS samples. Admissibility of the Bayes estimates are discussed and proved by the well-known Blyth’s lemma. Based on the available data, confidence intervals and highest posterior density credible sets are obtained. The generalized likelihood ratio (GLRT) and the Bayesian tests (under the “0 ? K” loss function) are derived for testing homogeneity of the exponential populations. It is shown that the GLRT in this case is scale invariant. Some guidelines for deriving the uniformly most powerful scale-invariant test (if exists) are also given.     

Nonnegative Unbiased Estimation of Scale Parameters and Associated Quantiles Based on a Ranked Set Sample

In this article, we are interested in estimating the scale parameter in location and scale families. It is well known that the best linear unbiased estimator (BLUE) of scale parameter based on a simple random sample (SRS) is nonnegative. However, the BLUE of scale parameter based on a ranked set sample (RSS) can assume negative values. We suggest various modifications of BLUE of scale parameter based on RSS so that the resulting estimators are unbiased as well as nonnegative. Their performances in terms of relative efficiencies are compared and some recommendations are made for normal, logistic, double exponential, two-parameter exponential and Weibull distributions. We also briefly discuss an application of the proposed nonnegative BLUE of scale parameter for quantile estimation for the above populations.     

Independence Structure of Natural Conjugate Densities to Exponential Families and the Gibbs' Sampler

In this paper the independence between a block of natural parameters and the complementary block of mean value parameters holding for densities which are natural conjugate to some regular exponential families is used to design in a convenient way a Gibbs' sampler with block updates. Even when the densities of interest are obtained by conditioning to zero a block of natural parameters in a density conjugate to a larger "saturated" model, the updates require only the computation of marginal distributions under the "unconditional" density. For exponential families which are closed under marginalization, including both the zero mean Gaussian family and the cross-classified Bernoulli family such an implementation of the Gibbs' sampler can be seen as an Iterative Proportional Fitting algorithm with random inputs.     

Bounds of moment generating functions of some life distributions

In this article we show that if a life has new better than used in expectation (NBUE) ageing property and if the mean life is finite then the moment generating function exists and is finite. In fact, the moment generating function is shown to be bounded above by that of the exponential distribution with the same mean. Analogous results are also proven for two much bigger families of life distribution, namely, the new better than renewal used in expectation (NBRUE) and the renewal new is better than used in expectation (RNBUE) and the renewal new better than renewal used in expectation (RNBRUE), provided that the life has finite two moments. Further, stronger results are also obtained for the smaller new better than used version of the above classes.     

Large deviations for estimators of some threshold parameters

In this paper, we consider a class of statistical models with a real-valued threshold parameter, which is either the minimum or the maximum of the support of the sampling distribution. We prove large deviation principles for sequences of estimators (maximum likelihood estimators and posterior distributions) as the sample size goes to infinity. Furthermore we illustrate some connections with the analogous large deviation results for the natural exponential families.     

Set Compound Estimation Under Entropy Loss

Set compound estimation has been studied for nearly half a century. This paper explores, for the first time, set compound estimation under entropy (Kullback-Leibler information) loss for a k -dimensional standard exponential family with a compact parameter space. It makes detailed investigation of entropy loss with the exponential family and related proper-ties. Asymptotically optimal set compound estimators with rates O ( n ^−1/2) under this loss are established for some discrete exponential families by using power series, representing the Bayes estimators in terms of a mixture density and applying the Singh-Datta Lemma. Poisson, negative binomial families and a two-dimensional model serve as examples.     

The Lack of Memory Property in the Density Form for the Bivariate Setup

The lack of memory property is a characterizing property of the exponential distribution in the continuous domain. In the bivariate setup different generalizations of the same are available in terms of survival function. We extend this lack of memory property in terms of bivariate probability density function and examine its characterization properties. In this process the density version of the lack of memory property can be interlinked with conditionally specified exponential distribution, bivariate reciprocal coordinate subtangent of the density curve and a few other derived measures.     

A TEST OF THE LOCATION PARAMETER OF THE EXPONENTIAL DISTRIBUTION

In this paper, we propose a procedure for testing the location parameter of the exponential distribution for certain alternative hypotheses, which could result in the early rejection of the null hypothesis. This is a consequence of the monotone property of the test statistic which is based on the extremal quotient. The test being scale-free does not require the scale parameter to be known.     

Estimation in the generalized linear empirical bayes model using the extended quasi-likelihood

A generalized linear empirical Bayes model is developed for empirical Bayes analysis of several means in natural exponential families. A unified approach is presented for all natural exponential families with quadratic variance functions (the Normal, Poisson, Binomial, Gamma, and two others.) The hyperparameters are estimated using the extended quasi-likelihood of Nelder and Pregibon (1987), which is easily implemented via the GLIM package. The accuracy of these estimates is developed by asymptotic approximation of the variance. Two data examples are illustrated.     

Analysis of rounded exponential data

The problem of inference based on a rounded random sample from the exponential distribution is treated. The main results are given by an explicit expression for the maximum-likelihood estimator, a confidence interval with a guaranteed level of confidence, and a conjugate class of distributions for Bayesian analysis. These results are exemplified on two concrete examples. The large and increasing body of results on the topic of grouped data has been mostly focused on the effect on the estimators. The methods and results for the derivation of confidence intervals here are hence of some general theoretical value as a model approach for other parametric models. The Bayesian credibility interval recommended in cases with a lack of other prior information follows by letting the prior equal the inverted exponential with a scale equal to one divided by the resolution. It is shown that this corresponds to the standard non-informative prior for the scale in the case of non-rounded data. For cases with the absence of explicit prior information it is argued that the inverted exponential prior with a scale given by the resolution is a reasonable choice for more general digitized scale families also.     

Une condition nécessaire d'admissibilité et ses conséquences sur les estimateurs à rétrécisseur de la moyenne d'un vecteur normal

Considering exponential families of distributions, we estimate parameters which are not the natural parameters. We prove that the admissible estimators of these parameters are limits of Bayes estimators and can be expressed through a given functional form. An important particular case of this model pertains to the estimation of the mean of a multidimensional normal distribution when the variance is known up to a multiplicative factor. We deduce from the main result a necessry condition for the admissibility of matricial shrinkage estimators.     

Density Approximation by Summary Statistics: An Information-theoretic Approach

In the case of exponential families, it is a straightforward matter to approximate a density function by use of summary statistics; however, an appropriate approach to such approximation is far less clear when an exponential family is not assumed. In this paper, a maximin argument based on information theory is used to derive a new approach to density approximation from summary statistics which is not restricted by the assumption of validity of an underlying exponential family. Information-theoretic criteria are developed to assess loss of predictive power of summary statistics under such minimal knowledge. Under these criteria, optimal density approximations in the maximin sense are obtained and shown to be related to exponential families. Conditions for existence of optimal density approximations are developed. Applications of the proposed approach are illustrated, and methods for estimation of densities are provided in the case of simple random sampling. Large-sample theory for estimates is developed.