Some Geometry of the Cone of Nonnegative Definite Matrices and Weights of Associated X2 Distribution

Consider the test problem about matrix normal mean M with the null hypothesis M = O against the alternative that M is nonnegative definite. In our previous paper (Kuriki (1993, Ann. Statist., 21, 1379–1384)), the null distribution of the likelihood ratio statistic has been given in the form of a finite mixture of ² distributions referred to as X² distribution. In this paper, we investigate differential-geometric structure such as second fundamental form and volume element of the boundary of the cone formed by real nonnegative definite matrices, and give a geometric derivation of this null distribution by virtue of the general theory on the X² distribution for piecewise smooth convex cone alternatives developed by Takemura and Kuriki (1997, Ann. Statist., 25, 2368–2387).     

A note on the uniformly most powerful tests in the presence of nuisance parameters

When a testing problem has nuisance parameters, the uniformly most powerful (UMP) tests do not generally exist. Exceptional examples were given by Dubey (1962, Skand. Aktuarietidskr., 45, 25–38; 1963, Skand. Aktuarietidskr., 46, 1–24) and Takeuchi (1968, Ann. Math. Statist., 40, 1838–1839) for the exponential distributions. What is essential for proving the existence of UMP tests lies in a special relationship between null hypothesis and the alternative. Assuming a similar relationship between them, a similar kind of result can be shown under more general situation.     

Likelihood ratio tests for comparingk populations —The two-parameter nonregular models

Summary The null and nonnull distributions of the likelihood ratio statistics for testing the homogeneity ofk given populations, each associated with a nonregular density depending on two truncation parameters, are investigated. This generalizes to the two-parameter case the work of Hogg (1956,Ann. Math. Statist.,27, 529–532), Barr (1966,J. Amer. Statist. Assoc.,61, 856–864) and Khatri and Jaiswal (1969,Aust. J. Statist.,11, 79–84; 1969, 1971,Ann. Inst. Statist. Math.,21, 127–136;23, 199–210).     

<Emphasis Type="Italic">I</Emphasis>-projection onto isotonic cones and its applications to maximum likelihood estimation for log-linear models

A frequently occurring problem is to find a probability vector,p∈D, which minimizes theI-divergence between it and a given probability vector π. This is referred to as theI-projection of π ontoD. Darroch and Ratcliff (1972,Ann. Math. Statist.,43, 1470–1480) gave an algorithm whenD is defined by some linear equalities and in this paper, for simplicity of exposition, we propose an iterative procedure whenD is defined by some linear inequalities. We also discuss the relationship betweenI-projection and the maximum likelihood estimation for multinomial distribution. All of the results can be applied to isotonic cone.     

Average worth and simultaneous estimation of the selected subset

Suppose a subset of populations is selected from the given k gamma G( _i,p ) (i = 1,2,...,k)populations, using Gupta's rule (1963, Ann. Inst. Statist. Math., 14, 199–216). The problem of estimating the average worth of the selected subset is first considered. The natural estimator is shown to be positively biased and the UMVUE is obtained using Robbins' UV method of estimation (1988, Statistical Decision Theory and Related Topics IV, Vol. 1 (eds. S. S. Gupta and J. O. Berger), 265–270, Springer, New York). A class of estimators that dominate the natural estimator for an arbitrary k is derived. Similar results are observed for the simultaneous estimation of the selected subset.     

Nonparametric test of restricted interchangeability

Exact and large sample distributions of the rank order test under the null hypothesis of restricted interchangeability are obtained. Under given regularity conditions and under Pitman's shift in location alternative, the asymptotic relative efficiency of this nonparametric test in comparison with Votaw's (1948, Ann. Math. Statist., 19, 447–473) likelihood ratio test is given.     

On Adaptive Combination of Regression Estimators

Consider the problem of choosing between two estimators of the regression function, where one estimator is based on stronger assumptions than the other and thus the rates of convergence are different. We propose a linear combination of the estimators where the weights are estimated by Mallows' C _L . The adaptive estimator retains the optimal rates of convergence and is an extension of Stein-type estimators considered by Li and Hwang (1984, Ann. Statist., 12, 887-897) and related to an estimator in Burman and Chaudhuri (1999, Ann. Inst. Statist. Math. (to appear)).     

A new Bayesian procedure for testing point null hypotheses

Testing point null hypotheses is a very common activity in various applied situations. However, the existing Bayesian testing procedure may give evidence which does not agree with the classical frequentist p-value in many point null testing situations. A typical example for this is the well known Lindley’s paradox (Lindley in Biometrika 44:187–192, 1957). In this paper we propose an alternative testing procedure in the Bayesian framework. It is shown that for many classical testing examples, the Bayesian evidence derived by our new testing procedure is not contradictory to its frequentist counterpart any more. In fact, the new Bayesian evidence under the noninformative prior is usually coincident with the frequentist observed significance level.     

Unbiased sequential estimation of1/p: Settlement of a conjecture

Summary We present a complete characterization of the class of (unbounded) sampling plans providing unbiased (sequential) estimation of the reciprocal of the Bernoulli parameterp. This settles a conjecture set forth by Sinha and Sinha (1975,Ann. Inst. Statist. Math.,27, 245–258) regarding the nature of such plans as sought out by Gupta (1967,Ann. Inst. Statist. Math.,19, 413–416). Incidentally, a special type of sampling plans (termed ‘infinite-step generalizations of the inverse binomial plans’), studied by Sinha and Bhattacharyya (1982, Institute of Statistics Mimeo Series, Raleigh), are seen to play a central role in this study.     

On the inadmissibility of preliminary-test estimators when the loss involves a complexity cost

Summary Estimation-preceded-by-testing is studied in the context of estimating the mean vector of a multivariate normal distribution under squared error loss together with a complexity cost. It is shown that although the preliminary test estimator is admissible for the univariate problem (cf Meeden and Arnold (1979),J. Amer. Statist. Assoc.,74, 872–874), for dimensionp≧3, the estimator is inadmissible. A new preliminary test estimator is obtained, which depends on the cost for each component and dominates the usual preliminary test estimator. Research partially supported by the NSF Grant Number DMS-82-18091 and partially by the DSR Research Development Award, University of Florida.     

Estimation of two normal means which may be common

Consider the problem of estimating the mean of a normal population when independent samples from this as well as a second normal population are available. Pre-test estimators which combine the two sample means if a test of the hypothesis of equal population means accepts but otherwise use only the first sample mean, are compared to limited translation estimators which are derived in the spirit of Bickel (1984, Ann. Statist., 12, 864–879) (we also cover the cases of unknown variances). Our conclusion is that if the accuracy with which the second population mean can be estimated is of the same or better order of magnitude as teh accuracy with which the first can be estimated, then the limited translation estimators largely dominate the pre-test estimators in terms of mean square error loss.This research was supported by grants from the FRD of the CSIR of South Africa.     

Estimation of the ratio of the scale parameters of two exponential distributions with unknown location parameters

We consider the estimation of the ratio of the scale parameters of two independent two-parameter exponential distributions with unknown location parameters. It is shown that the best affine equivariant estimator (BAEE) is inadmissible under any loss function from a large class of bowl-shaped loss functions. Two new classes of improved estimators are obtained. Some values of the risk functions of the BAEE and two improved estimators are evaluated for two particular loss functions. Our results are parallel to those of Zidek (1973, Ann. Statist., 1, 264–278), who derived a class of estimators that dominate the BAEE of the scale parameter of a two-parameter exponential distribution.     

On Aki's nonparametric test for symmetry

Summary Aki (1987,Ann. Inst. Statist. Math,39, 457–472) develops a theory of extending the test for symmetry about zero of a continuous distribution functionF. In this paper we discuss the same testing problem in the case where the probabilityF(0) of being negative is unknown, which is assumed to be known in Aki's paper.     

On the distribution of the total number of run lengths

In the present paper, we study the distribution of a statistic utilizing the runs length of “reasonably long” series of alike elements (success runs) in a sequence of binary trials. More specifically, we are looking at the sum of exact lengths of subsequences (strings) consisting ofk or more consecutive successes (k is a given positive integer). The investigation of the statistic of interest is accomplished by exploiting an appropriate generalization of the Markov chain embedding technique introduced by Fu and Koutras (1994,J. Amer. Statist. Assoc.,89, 1050–1058) and Koutras and Alexandrou (1995,Ann. Inst. Statist. Math.,47, 743–766). In addition, we explore the conditional distribution of the same statistic, given the number of successes and establish statistical tests for the detection of the null hypothesis of randomness versus the alternative hypothesis of systematic clustering of successes in a sequence of binary outcomes. Research supported by General Secretary of Research and Technology of Greece under grand PENED 2001.     

Response surface models for the Leybourne unit root tests and lag order dependence

This paper calculates response surface models for a large range of quantiles of the Leybourne (Oxf Bull Econ Stat 57:559–571, 1995) test for the null hypothesis of a unit root against the alternative of (trend) stationarity. The response surface models allow the estimation of critical values for different combinations of number of observations, T, and lag order in the test regressions, p, where the latter can be either specified by the user or optimally selected using a data-dependent procedure. The results indicate that the critical values depend on the method used to select the number of lags. An Excel spreadsheet is available to calculate the p-value associated with a test statistic.     

The quantum generalization of optimal statistical estimation and hypothesis testing

A review of the fundamental ideas and methods of the optimal reception and processing of quantum signals is given. Estimation via an operator based on the usual and generalized measurements (e.g., quasi-measurements) are discussed. The theory of operator estimation enables one to obtain Bayes limits for the usual estimates. The change from usual measurements to quasi-measurements generally improves performance.

We show that some quasi-measurements can be realized as indirect measurements, and introduce an operator measure II(db) which describes the quasi-measurements on the space of measurements.

Estimation based on quasi-measurements is described by operator measures Q(du) on the space of estimates. Minimization with respect to Q(du) is minimization simultaneously for all quasi-measurements and all estimates.

For the M-ary hypothesis testing problem finding the optimum reduces to finding non-negative definite Hermitian operators Q₁,…, Q _m satisfying Q₁+…+Q _m = 1 where 1 is the identity operator, which extremize.

Optimal Bayes quantum estimation is discussed. In the case of Gaussian quantum signal and minimum variance estimation, finding the quasi-measurements leads to optimal linear estimation. Further suboptimal methods for finding the operator in more general cases are discussed.     

r-k Class estimation in regression model with concomitant variables

We treat with the r-k class estimation in a regression model, which includes the ordinary least squares estimator, the ordinary ridge regression estimator and the principal component regression estimator as special cases of the r-k class estimator. Many papers compared total mean square error of these estimators. Sarkar (1989, Ann. Inst. Statist. Math., 41, 717–724) asserts that the results of this comparison are still valid in a misspecified linear model. We point out some confusions of Sarkar and show additional conditions under which his assertion holds.     

Improved sequential estimation of means of exponential distributions

Both one-sample and multi-sample estimation problems for the means of one parameter exponential distributions are addressed. In the one-sample case, for the existing purely sequential and recently obtained piecewise sequential estimation methodologies, we follow and extend the development in Isogai and Uno (1993,Ann. Inst. Statist. Math. (in press)) in order to obtain a class of estimators that provides asymptotic second-order risk improvement. In the multi-sample problem, we address the analogous aspects for the existing purely sequential methodology as well as the newly developed piecewise methodology.     

Bivariate Sign Tests Based on the Sup, L 1 and L 2 Norms

The bivariate location problem is considered. The sup, L ₁ and L ₂ norms are used to construct bivariate sign tests from the univariate sign statistics computed on the projected observations on all lines passing through the origin. The tests so obtained are affine-invariant and distribution-free under the null hypothesis. The sup-norm gives rise to Hodges' test. A class of tests derived from the L ₂-norm, with Blumen's test as a member, is seen to be related to a class proposed by Oja and Nyblom (1989, J. Amer. Statist. Assoc., 84, 249-259). The L ₁-norm gives rise to a new test. Its asymptotic null distribution is seen to be the same as that of the L ₁-norm of a certain normal process related to the standard Wiener process. An explicit expression of its cumulative distribution function is given. A simulation study will examine the merits of the three approaches.     

Positron Emission Tomography and Random Coefficients Regression

We further explore the relation between random coefficients regression (RCR) and computerized tomography. Recently, Beran et al. (1996, Ann. Statist., 24, 2569–2592) explored this connection to derive an estimation method for the non-parametric RCR problem which is closely related to image reconstruction methods in X-ray computerized tomography. In this paper we emphasize the close connection of the RCR problem with positron emission tomography (PET). Specifically, we show that the RCR problem can be viewed as an idealized (continuous) version of a PET experiment, by demonstrating that the nonparametric likelihood of the RCR problem is equivalent to that of a specific PET experiment. Consequently, methods independently developed for either of the two problems can be adapted from one problem to the other. To demonstrate the close relation between the two problems we use the estimation method of Beran, Feuerverger and Hall for image reconstruction in PET.