Contre-exemple au théorème de P.C. Consul sur la factorisation des distributions de Poisson généralisées

This note exhibits two independent random variables on integers, X₁ and X₂, such that neither X₁ nor X₂ has a generalized Poisson distribution, but X₁ + X₂ has. This contradicts statements made by Professor Consul in his recent book.     

Correction locale de l'estimateur à noyau de la densité d'une loi de probabilité

The standard Parzen-Rosenblatt kernel density estimator is known to systematically deviate from the true value near critical points of the density curve. To overcome this difficulty, we extend the Rao-Blackwell method by using locally sufficient statistics: we define a new estimator and study its asymptotic behaviour. The interest of the method is shown by means of simulations.     

Sur la linéarité asymptotique du premier ordre de statistiques de rang signé

Two first-order uniform asymptotic linearity theorems for signed-rank statistics are given which generalize similar theorems of Jure?ková [Sankhyā Ser. A, 33, 1-18 (1971)], van Eeden [Ann. Math. Statist., 43, 791-802 (1972)], and Kraft and van Eeden [Ann. Math. Statist., 43, 42-57 (1972)]. It is seen that the concordance conditions imposed by these authors are not needed.     

Taux de résistance des tests de rang d'indépendance

The resistance of tests to acceptance and rejection of null hypotheses was denned and studied by Ylvisaker in the context of one-sample problems. This notion provides a measure of a test's resistance to outliers. In this paper, we propose an extension of this notion to rank-based tests of independence for bivariate random variables. We show, among other things, that Kendall's test of independence is more resistant than Spearman's test.     

Estimation suroptimale de la densité par projection

Superefficiency of a projection density estimator The author constructs a projection density estimator with a data‐driven truncation index. This estimator reaches the superoptimal rates 1/n in mean integrated square error and {In ln(n/n}^1/2 in uniform almost sure convergence over a given subspace which is dense in the class of all possible densities; the rate of the estimator is quasi‐optimal everywhere else. The subspace in question may be chosen a priori by the statistician.     

De l'unicité des estimateurs robustes en régression lorsque le paramètre d'échelle et le paramètre de la régression sont estimés simultanément

Several methods have been suggested to calculate robust M- and G-M -estimators of the regression parameter β and of the error scale parameter σ in a linear model. This paper shows that, for some data sets well known in robust statistics, the nonlinear systems of equations for the simultaneous estimation of β, with an M-estimate with a redescending ψ-function, and σ, with the residual median absolute deviation (MAD), have many solutions. This multiplicity is not caused by the possible lack of uniqueness, for redescending ψ-functions, of the solutions of the system defining β with known σ; rather, the simultaneous estimation of β and σ together creates the problem. A way to avoid these multiple solutions is to proceed in two steps. First take σ as the median absolute deviation of the residuals for a uniquely defined robust M-estimate such as Huber's Proposal 2 or the L₁-estimate. Then solve the nonlinear system for the M-estimate with σ equal to the value obtained at the first step to get the estimate of β. Analytical conditions for the uniqueness of M and G-M-estimates are also given.     

Propriétés statistiques des copules de valeurs extrêmes bidimensionnelles

Let (X, Y) be a bivariate random vector whose distribution function H(x, y) belongs to the class of bivariate extreme-value distributions. If F₁ and F₂ are the marginals of X and Y, then H(x, y) = C{F₁(x),F₂(y)}, where C is a bivariate extreme-value dependence function. This paper gives the joint distribution of the random variables Z = {log F₁(X)}/{log F₁(X)F₂(Y)} and W = C{F₁{(X),F₂(Y)}. Using this distribution, an algorithm to generate random variables having bivariate extreme-value distribution is présentés. Furthermore, it is shown that for any bivariate extreme-value dependence function C, the distribution of the random variable W = C{F₁(X),F₂(Y)} belongs to a monoparametric family of distributions. This property is used to derive goodness-of-fit statistics to determine whether a copula belongs to an extreme-value family.     

L'analyse de la variation saisonniére quand l'amplitude et la taille sont faibles

Studies of seasonal variation are valuable in biomedical research because they can help to discover the etiology of diseases that are not well understood. Generally in these studies the data have certain characteristics that require specialized tests and methods for the statistical analysis. But the effectiveness of these specialized tests is variable, especially according to the seasonal variation, the dimension of the amplitude in the seasonal variation, and the sample size. The purpose of this paper is to present a test and methods appropriate for the analysis and modeling of data whose seasonal variation has small amplitude and whose sample size is small. This test can detect different kinds of seasonal variation. The results from a simulation study show that the test performs very well. The application of these methods is illustrated by two examples.     

Une application des méthodes séquentielles aux tests de permutations

To carry out a permutation test we have to examine the n! permutations of the observations. In order to make the permutation test feasible, Dwass (1957) proposed to examine only a sample of these permutations. With the help of sequential methods, we obtain a test which is never less efficient than that proposed by Dwass or the permutation test itself, in the sense that it is as powerful and never requires more permutations to make a decision. In practice, we can expect to gain much efficiency.     

Sur la convergence des tests de Schlee et de Yatchew

L'auteur adapte l'approche de Schlee (1980) et de Yatchew (1992) à la construction de deux tests de convexité pour une fonction de régression dans un modèle non paramétrique. Il établit de nouveaux résultats de convergence pour ces tests et étudie leur comportement dans de petits échantillons à l'aide de simulations.     

Réduction de la variance dans les sondages en présence d'information auxiliarie: Une approache non paramétrique par splines de régression

The author considers the use of auxiliary information available at population level to improve the estimation of finite population totals. She introduces a new type of model‐assisted estimator based on nonparametric regression splines. The estimator is a weighted linear combination of the study variable with weights calibrated to the B‐splines known population totals. The author shows that the estimator is asymptotically design‐unbiased and consistent under conditions which do not require the superpopulation model to be correct. She proposes a design‐based variance approximation and shows that the anticipated variance is asymptotically equivalent to the Godambe‐Joshi lower bound. She also shows through simulations that the estimator has good properties.     

L'effet de l'erreur d'arrondi sur l'estimateur d'une densité par la méthode du noyau

In a model for rounded data suppose that the random sample X₁,.,.,X_n,. i.i.d., is transformed into an observed random sample X_1Δ,.,.,X_nΔ, where X_iΔ = 2vΔ if X_i, ∈ (2vΔ - Δ, 2vΔ + Δ), for i = 1,.,.,n. We show that the precision Δ of the observations has an important effect on the shape of the kernel density estimator, and we identify important points for the graphical display of this estimator. We examine the IMSE criteria to find the optimal window under the rounded-data model.     

Test de différence de contrastes et somme pondérée de khi-deux

After recalling the framework of minimum-contrast estimation, its consistency and its asymptotic normality, we highlight the fact that these results do not require any stationarity or ergodicity assumptions. The asymptotic distribution of the underlying contrast difference test is a weighted sum of independent chi-square variables having one degree of freedom each. We illustrate these results in three contexts: (1) a nonhomogeneous Markov chain with likelihood contrast; (2) a Markov field with coding, pseudolikelihood or likelihood contrasts; (3) a not necessarily Gaussian time series with Whittle's contrast. In contexts (2) and (3), we compare experimentally the power of the likelihood-ratio test with those of other contrast-difference tests.     

Approximation spline de la prevision d'un processus fonctionnel autorégressif d'ordre 1

ABSTRACT We present a method to approximate and forecast, on an entire interval, a continuous-time process. For this purpose, we use the modelization of ARH(l) processes, defined by Bosq (1991). We deal with the practical problem of the discretization of the observed trajectories and approximate them by means of spline functions. We show by simulations that for well-chosen smoothing parameters, good prediction can be obtained in comparison with the “predictable” part of the process. Finally, we apply this model to forecast road traffic and compare it with a SARIMA model.     

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.     

Concepts de dépendance et ordres stochastiques pour des lois bidimensionnelles

Yanagimoto and Okamoto (1969) introduced a stochastic ordering that generalizes a concept of monotone regression dependence introduced by Lehmann (1966). In this paper, we define and examine the properties of three new orderings which imply that of Yanagimoto and Okamoto. One of these orderings is seen to extend Shaked's (1977) notion of DTP(0, 1), and another includes Lehmann's concept of positive likelihood-ratio dependence as a special case. The proposed orderings are also compared with the TP₂ positive-dependence ordering defined by Kimeldorf and Sampson (1987).     

Estimation du comportement asymptotique des autocovariances et autocorrelations empiriques de processus multivariéeas

Statistics based on the sample autocovariances are widely used in time-series analysis. Estimators of the asymptotic covariance between the sample autocovariances are commonly derived from the so-called Bartlett's formula. However, this formula essentially holds for linear processes. This entails that for a wide range of nonlinear time series the above-mentioned estimators are not suitable. In this paper the behaviour of an alternative estimator is studied within the framework of centered or uncentered multivariate strongly mixing processes. Applications to differential functions of sample autocovariances, such as the sample autocorrelations, are considered.