Testing for and against a set of linear inequality constraints in a multinomial setting

There are numerous situations in categorical data analysis where one wishes to test hypotheses involving a set of linear inequality constraints placed upon the cell probabilities. For example, it may be of interest to test for symmetry in k × k contingency tables against one-sided alternatives. In this case, the null hypothesis imposes a set of linear equalities on the cell probabilities (namely p_ij = P_ji ×i > j), whereas the alternative specifies directional inequalities. Another important application (Robertson, Wright, and Dykstra 1988) is testing for or against stochastic ordering between the marginals of a k × k contingency table when the variables are ordinal and independence holds. Here we extend existing likelihood-ratio results to cover more general situations. To be specific, we consider testing H_t,0 against H₁ - H₀ and H₁ against H₂ - H ₁ when H₀:k × i=1 p_ix_ji = 0, j = 1,…, s, H₁:k × i=1 p_ix_ji × 0, j = 1,…, s, and does not impose any restrictions on p. The x_ji's are known constants, and s × k - 1. We show that the asymptotic distributions of the likelihood-ratio tests are of chi-bar-square type, and provide expressions for the weighting values.     

Conditional Tests for censored matched pairs

A class of statistics is proposed for the problem of testing for location difference using randomly right censored matched pair data. Each member of the class provides a conditionally distribution-free test of H₀: no location difference. Simulation results indicate that powers tests basd on certain members in the class are as good as or better thatn the power of a test proposed by Woolson and Lachenbruch (1980).     

Optimal linear combinations using householder transformations

In pattern classification of sampled vector valued random variables it is often essential, due to computational and accuracy considerations, to consider certain measurable transformations of the random variable. These transformations are generally of a dimension-reducing nature. In this paper we consider the class of linear dimension reducing transformations, i.e., the k × n matrices of rank k where k < n and n is the dimension of the range of the sampled vector random variable.

In this connection, we use certain results (Decell and Quirein, 1973), that guarantee, relative to various class separability criteria, the existence of an extremal transformation. These results also guarantee that the extremal transformation can be expressed in the form (I_k∣ Z)U where I_k is the k × k identity matrix and U is an orthogonal n × n matrix. These results actually limit the search for the extremal linear transformation to a search over the obviously smaller class of k × n matrices of the form (I_k ∣Z)U. In this paper these results are refined in the sense that any extremal transformation can be expressed in the form (I_K∣Z)H_p … H₁ where p ≤ min{k, n?k} and H_i is a Householder transformation i=l,…, p, The latter result allows one to construct a sequence of transformations (L_K∣ Z)H₁, (I_K Z)H₂H₁ … such that the values of the class separability criterion evaluated at this sequence is a bounded, monotone sequence of real numbers. The construction of the i-th element of the sequence of transformations requires the solution of an n-dimensional optimization problem. The solution, for various class separability criteria, of the optimization problem will be the subject of later papers. We have conjectured (with supporting theorems and empirical results) that, since the bounded monotone sequence of real class separability values converges to its least upper bound, this least upper bound is an extremal value of the class separability criterion.

Several open questions are stated and the practical implications of the results are discussed.     

Uniform hypothesis testing for finite-valued stationary processes

Given a discrete-valued sample X₁, …, X_n, we wish to decide whether it was generated by a distribution belonging to a family H₀, or it was generated by a distribution belonging to a family H₁. In this work, we assume that all distributions are stationary ergodic, and do not make any further assumptions (e.g. no independence or mixing rate assumptions). We would like to have a test whose probability of error (both Types I and II) is uniformly bounded. More precisely, we require that for each ? there exists a sample size n such that probability of error is upper-bounded by ? for samples longer than n. We find some necessary and some sufficient conditions on H₀ and H₁ under which a consistent test (with this notion of consistency) exists. These conditions are topological, with respect to the topology of distributional distance.     

Location adjustment for the minimum volume ellipsoid estimator

Estimating multivariate location and scatter with both affine equivariance and positive breakdown has always been difficult. A well-known estimator which satisfies both properties is the Minimum Volume Ellipsoid Estimator (MVE). Computing the exact MVE is often not feasible, so one usually resorts to an approximate algorithm. In the regression setup, algorithms for positive-breakdown estimators like Least Median of Squares typically recompute the intercept at each step, to improve the result. This approach is called intercept adjustment. In this paper we show that a similar technique, called location adjustment, can be applied to the MVE. For this purpose we use the Minimum Volume Ball (MVB), in order to lower the MVE objective function. An exact algorithm for calculating the MVB is presented. As an alternative to MVB location adjustment we propose L ₁ location adjustment, which does not necessarily lower the MVE objective function but yields more efficient estimates for the location part. Simulations compare the two types of location adjustment. We also obtain the maxbias curves of L ₁ and the MVB in the multivariate setting, revealing the superiority of L ₁.     

On comparing two poisson intensity functions

Given realizations of two possion processes with unknown intensities A(·) and F(·) observed over the interval (t₁,t₂), we suppose that it is desired to distinution between H₀ Ξ(·)/λ(·) is constant on (t₁,t₂) versus H₊:Ξ(·)/λ(·) increases on (t₁,t₂). We propose a decision rule which uses the percentage points of the Mann-Whitney U-distribution. We show that the decision rule is unbiased and that the set of alternatives in H₊ can be weakly ordered, specifically: if Ξ(·)/λ(·), β(·)/λ(·) and Ξ(·)/β(·) are increasing on (t₁, t₂) then P{H₀ is rejected |Ξ(·)}≧P{H₀ is rejected|B(·)}≧P{H₀ is rejected|H₀}.     

A combined p-value test for multiple hypothesis testing

Tests that combine p-values, such as Fisher's product test, are popular to test the global null hypothesis H₀ that each of n component null hypotheses, H₁,…,H_n, is true versus the alternative that at least one of H₁,…,H_n is false, since they are more powerful than classical multiple tests such as the Bonferroni test and the Simes tests. Recent modifications of Fisher's product test, popular in the analysis of large scale genetic studies include the truncated product method (TPM) of Zaykin et al. (2002), the rank truncated product (RTP) test of Dudbridge and Koeleman (2003) and more recently, a permutation based test—the adaptive rank truncated product (ARTP) method of Yu et al. (2009). The TPM and RTP methods require users' specification of a truncation point. The ARTP method improves the performance of the RTP method by optimizing selection of the truncation point over a set of pre-specified candidate points. In this paper we extend the ARTP by proposing to use all the possible truncation points {1,…,n} as the candidate truncation points. Furthermore, we derive the theoretical probability distribution of the test statistic under the global null hypothesis H₀. Simulations are conducted to compare the performance of the proposed test with the Bonferroni test, the Simes test, the RTP test, and Fisher's product test. The simulation results show that the proposed test has higher power than the Bonferroni test and the Simes test, as well as the RTP method. It is also significantly more powerful than Fisher's product test when the number of truly false hypotheses is small relative to the total number of hypotheses, and has comparable power to Fisher's product test otherwise.     

Kolmogorov-type tests of goodness-of-fit against stochastic ordering

This article addresses the problem of testing the null hypothesis H₀ that a random sample of size n is from a distribution with the completely specified continuous cumulative distribution function F_n(x). Kolmogorov-type tests for H₀ are based on the statistics C⁺ _n = Sup[F_n(x)?F₀(x)] and C^? _n=Sup[F₀(x)?F_n(x)], where F_n(x) is an empirical distribution function. Let F(x) be the true cumulative distribution function, and consider the ordered alternative H₁: F(x)≥F₀(x) for all x and with strict inequality for some x. Although it is natural to reject H₀ and accept H₁ if C ⁺ _n is large, this article shows that a test that is superior in some ways rejects F₀ and accepts H₁ if C^mdash _n is small. Properties of the two tests are compared based on theoretical results and simulated results.     

Maximum term of a particular autoregressivesequence with discrete margins

In this paper we consider a stationary sequence of discrete random variables with marginal distribution H(x), obtained by a simple transformation from the max-AR(1) sequence considered by Alpuim (1989). Because discrete distributions impose severe restrictions on the convergence of the normalized maxima to an extreme value distribution, it is seen that in this particular case, whenever H(x) belongs to the domain of attraction of any max-stable distribution, the sequence possesses an extremal index 0 = 0. Nevertheless, it, is possible to obtain a nondegenerate limiting distribution for the linearized maxima by choosing other sets of normalizing constants. Whenever H(x) does not belong to the domain of attraction of any max-stable distribution, but, satisfies adequate conditions, the maxima nearly possess an asymptotic stability with the presence of an extremal index 0 <θ<1.

Motivated by the behaviour of these sequences we obtained a more general result extending the results of Anderson (1970) and Me (Jon nick and Park (1992) over the mixing conditionsD ^(k)(u_n), defined by Chermck et al (1991).

Several examples, obtained after simulation, are presented in order to illustrate the different situations that may occur.     

A simple algorithm for the adaptation of scores and power behavior of the corresponding rank test

The purpose of this paper is twofold:On one hand we want to give a very simple algorithm for evaluating a special rank estimator of the type given in Behnen, Neuhaus, and Ruymgaart (1983) for the approximate optimal choice of the scores-generating function of a two-sample linear rank test for the general testing problem H_o:F=G versus H₁:F ≤ G, F ≠ G, in order to demonstrate that the corresponding adaptive rank statistic is simple enough for practical applications. On the other hand we prove the asymptotic normality of the adaptive rank statistic under H (leading to approximate critical values) and demonstrate the adaptive behavior of the corresponding rank test by a Monte Carlo power simulation for sample sizes as low as m=10, n=10.     

Discrimination analysis when the variates are grouped and observed in sequential order

Suppose that measurements Math', i = l,....,k, are consecutively taken on an individual at the prescribed costs C_i, i = l,....,k. the individual comes from one of the two populations H₁ and H₂, and it is desired to detect which population the individual belongs to. Given the loss incurreed in selecting population H_r when in fact it belongs to H_s, the prior probability P_r of H_r (r = 1,2), and assuming that H_r has the normal distribution N(µ_r, V), r = 1,2, we derive the sequential Bayesian solution of the discrimination problem when µ₁, µ₂ and V are known. When µ_r, V are unknown and must be estimated, we propose a solution which is asymptotic Bayesian with exponential convergence rate.     

Fast Poisson noise removal by biorthogonal Haar domain hypothesis testing

Methods based on hypothesis tests (HTs) in the Haar domain are widely used to denoise Poisson count data. Facing large datasets or real-time applications, Haar-based denoisers have to use the decimated transform to meet limited-memory or computation-time constraints. Unfortunately, for regular underlying intensities, decimation yields discontinuous estimates and strong “staircase” artifacts. In this paper, we propose to combine the HT framework with the decimated biorthogonal Haar (Bi-Haar) transform instead of the classical Haar. The Bi-Haar filter bank is normalized such that the p-values of Bi-Haar coefficients (p_BH) provide good approximation to those of Haar (p_H) for high-intensity settings or large scales; for low-intensity settings and small scales, we show that p_BH are essentially upper-bounded by p_H. Thus, we may apply the Haar-based HTs to Bi-Haar coefficients to control a prefixed false positive rate. By doing so, we benefit from the regular Bi-Haar filter bank to gain a smooth estimate while always maintaining a low computational complexity. A Fisher-approximation-based threshold implementing the HTs is also established. The efficiency of this method is illustrated on an example of hyperspectral-source-flux estimation.     

Tests based on equivariant estimators

Let ?⁽¹⁾ and ?⁽²⁾ be location-equivariant estimators of an unknown location parameter μ. It is shown that the test for H₀: μ ≤ μ₀ versus H_A : μ > μ₀ that rejects H₀ if ?⁽¹⁾ is large is uniformly more powerful than the one that rejects H₀ if ?⁽²⁾ is large if and only if ?⁽²⁾ is “more dispersed” than ?⁽¹⁾. A similar result is obtained for tests on scale using the star-shaped ordering. Examples are given.     

Testing for markov process vs semi-markov process

Here we have obtained a non-parametric test for testing the null hypothesis H₀ that the given realization is from a Markov process against the alternative hypothesis H₁ that it is from a semi-Markov process with the transition rate monotone increasing (or decreasing). We have shown that the test criterion has normal distribution asymptotically, and the test is consistent and unbiased.     

Asymptotic Distribution for Symmetric Spacing Statistics

A new procedure for testing the H ₀: μ₁ = ··· = μ _k against the alternative H _u :μ₁ ≥ ··· ≥μ _r  ≤ ··· ≤ μ _k with at least one strict inequality, where μ _i is the location parameter of the ith two-parameter exponential distribution, i = 1,…, k, is proposed. Exact critical constants are computed using a recursive integration algorithm. Tables containing these critical constants are provided to facilitate the implementation of the proposed test procedure. Simultaneous confidence intervals for certain contrasts of the location parameters are derived by inverting the proposed test statistic. In comparison to existing tests, it is shown, by a simulation study, that the new test statistic is more powerful in detecting U-shaped alternatives when the samples are derived from exponential distributions. As an extension, the use of the critical constants for comparing Pareto distribution parameters is discussed.     

Estimation of the location parameter and the average worth of the selected subset of two parameter exponential populations under LINEX loss function

Suppose a subset of populations is selected from k exponential populations with unknown location parameters θ₁, θ₂, …, θ_k and common known scale parameter σ. We consider the estimation of the location parameter of the selected population and the average worth of the selected subset under an asymmetric LINEX loss function. We show that the natural estimator of these parameters is biased and find the uniformly minimum risk-unbiased (UMRU) estimator of these parameters. In the case of k = 2, we find the minimax estimator of the location parameter of the smallest selected population. Furthermore, we compare numerically the risk of UMRU, minimax, and the natural estimators.     

Minimum hellinger distance estimation for normal models

A robust estimator introduced by Beran (1977a, 1977b), which is based on the minimum Hellinger distance between a projection model density and a nonparametric sample density, is studied empirically. An extensive simulation provides an estimate of the small sample distribution and supplies empirical evidence of the estimator performance for a normal location-scale model. While the performance of the minimum Hellinger distance estimator is seen to be competitive with the maximum likelihood estimator at the true model, its robustness to deviations from normality is shown to be competitive in this setting with that obtained from the M-estimator and the Cramér-von Mises minimum distance estimator. Beran also introduced a goodness-of-fit statisticH ₂, based on the minimized Hellinger distance between a member of a parametric family of densities and a nonparametric density estimate. We investigate the statistic H (the square root of H ₂) as a test for normality when both location and scale are unspecified. Empirically derived critical values are given which do not require extensive tables. The power of the statistic H compares favorably with four other widely used tests for normality.     

Estimators of location based on Kolmogorov-Smirnov-type statistics

Two classes of estimators of a location parameter ø₀ are proposed, based on a nonnegative functional H₁^* of the pair (D₁^ø_N, G^ø_N), where and where F_N is the sample distribution function. The estimators of the first class are defined as a value of ø minimizing H₁^*; the estimators of the second class are linearized versions of those of the first. The asymptotic distribution of the estimators is derived, and it is shown that the Kolmogorov-Smirnov statistic, the signed linear rank statistics, and the Cramérvon Mises statistics are special cases of such functionals H₁^*;. These estimators are closely related to the estimators of a shift in the two-sample case, proposed and studied by Boulanger in B2 (pp. 271–284).     

Projector operators in the multivariate Zyskind-Martin model

Summary Two quadratic formsS _H andS _E for a testable hypothesis and for an error in the multivariate Zyskind-Martin model with singular covariance matrix are expressed by means of projector operators. Thus the results for the multivariate standard model with identity covariance matrix given by Humak (1977) and Christensen (1987, 1991) are generalized for the case of Zyskind-Martin model. Special cases of our results are formulae forS _H andS _E in Aitken's (1935) model. In the case of general Gauss-Markoff modelS _H andS _E can also be expressed by means of projector operators for some subclasses of testable hypotheses. For these hypotheses, testing in Gauss-Markoff model is equivalent to testing in a Zyskind-Martin model.     

COMPARISON OF THREE NON-PARAMETRIC TESTS FOR BIVARIATE INTERCHANGEABILITY1

Given a random sample(X₁, Y₁), …,(X_n, Y_n) from a bivariate (BV) absolutely continuous c.d.f. H (x, y), we consider rank tests for the null hypothesis of interchangeability H₀: H(x, y). Three linear rank test statistics, Wilcoxon (W_N), sum of squared ranks (SSR_N) and Savage (S_N), are described in Section 1. In Section 2, asymptotic relative efficiency (ARE) comparisons of the three types of tests are made for Morgenstern (Plackett, 1965) and Moran (1969)BV alternatives with marginal distributions satisfying G(x) = F(x/θ) for some θ≠ 1. Both gamma and lognormal marginal distributions are used.