Gilles Ducharme
Université de Montpellier, Mathématiques, Faculty Member
Research Interests:
Research Interests:
In regression analysis, a best subset of regressors is usually selected by minimizing Mallows's C statistic or some other equivalent criterion, such as the Akaike lambda information criterion or cross-validation. It is known that the... more
In regression analysis, a best subset of regressors is usually selected by minimizing Mallows's C statistic or some other equivalent criterion, such as the Akaike lambda information criterion or cross-validation. It is known that the resulting procedure suffers from a lack of consistency that can lead to a model with too many variables. For this reason, corrections have been proposed that yield consistent procedures. The object of this paper is to show that these corrected criteria, although asymptotically consistent, are usually too conservative for finite sample sizes. The paper also proposes a new correction of Mallows's statistic that yields better results. A simulation study is conducted that shows that the proposed criterion performs well in a variety of situations.
Research Interests:
Research Interests:
... B (1986) 48, No. 2, pp. 197-205 Testing Collapsibility in Contingency Tables By GILLES R. DUCHARME and YVES LEPAGE Universite de Montreal, Canada [Received February 1985. Revised February 1986] SUMMARY In this article, the problem of... more
... B (1986) 48, No. 2, pp. 197-205 Testing Collapsibility in Contingency Tables By GILLES R. DUCHARME and YVES LEPAGE Universite de Montreal, Canada [Received February 1985. Revised February 1986] SUMMARY In this article, the problem of testing for the collapsibility ...
Research Interests:
Research Interests:
Research Interests:
ABSTRACT The asymptotic distribution of the circular median is derived for symmetric distributions on the circle. Its asymptotic relative efficiency with respect to the mean direction and to an stimator proposed by G. S. Watson... more
ABSTRACT The asymptotic distribution of the circular median is derived for symmetric distributions on the circle. Its asymptotic relative efficiency with respect to the mean direction and to an stimator proposed by G. S. Watson [Statistics on spheres (1983; Zbl 0646.62045)] is then examined. Special attention is given to the cases where the underlying distribution is von Mises and contaminated von Mises. It is seen that the circular median can perform more efficiently than both estimators in presence of outliers.
Research Interests:
Research Interests:
In this paper, tests of goodness-of-fit for the inverse Gaussian distribution are developed. The distribution involves a shape parameter and, because of this, some test approaches lead to inconsistent strategies. A consistent test is... more
In this paper, tests of goodness-of-fit for the inverse Gaussian distribution are developed. The distribution involves a shape parameter and, because of this, some test approaches lead to inconsistent strategies. A consistent test is proposed and its properties investigated. A table of critical points is provided and both the level and the power of the test are explored by simulation.
Research Interests:
ABSTRACT A measure of multivariate correlation between two sets of vectors is considered when the underlying joint distribution is a member of the class of elliptical distributions. Its asymptotic distribution is derived under different... more
ABSTRACT A measure of multivariate correlation between two sets of vectors is considered when the underlying joint distribution is a member of the class of elliptical distributions. Its asymptotic distribution is derived under different situations and these results are used to test hypotheses on vector correlation when the underlying joint distribution is non-normal.
Research Interests:
Research Interests:
Bacro Jean-Noël Président du Jury Berlinet Alain Examinateur Ducharme Gilles Directeur de Thèse Harel Michel Rapporteur Saracco Jérôme Examinateur Sarda Pascal Rapporteur
Research Interests:
Research Interests:
In this paper, tests of goodness-of-fit for the inverse Gaussian distribution are developed. The distribution involves a shape parameter and, because of this, some test approaches lead to inconsistent strategies. A consistent test is... more
In this paper, tests of goodness-of-fit for the inverse Gaussian distribution are developed. The distribution involves a shape parameter and, because of this, some test approaches lead to inconsistent strategies. A consistent test is proposed and its properties investigated. A table of critical points is provided and both the level and the power of the test are explored by simulation.
Research Interests:
Research Interests:
Research Interests:
Research Interests:
In this paper we develop, for directional and axial data, smooth tests of goodness-of-fit for rotationally symmetric distributions against general families of embedding alternatives constructed from complete orthonormal bases of... more
In this paper we develop, for directional and axial data, smooth tests of goodness-of-fit for rotationally symmetric distributions against general families of embedding alternatives constructed from complete orthonormal bases of functions. These families generalize a proposal of Beran (1979) ...
Research Interests:
In regression analysis, a best subset of regressors is usually selected by minimizing Mallows's C statistic or some other equivalent criterion, such as the Akaike lambda information criterion or cross-validation. It is known that the... more
In regression analysis, a best subset of regressors is usually selected by minimizing Mallows's C statistic or some other equivalent criterion, such as the Akaike lambda information criterion or cross-validation. It is known that the resulting procedure suffers from a lack of consistency that can lead to a model with too many variables. For this reason, corrections have been proposed that yield consistent procedures. The object of this paper is to show that these corrected criteria, although asymptotically consistent, are usually too conservative for finite sample sizes. The paper also proposes a new correction of Mallows's statistic that yields better results. A simulation study is conducted that shows that the proposed criterion performs well in a variety of situations.
Research Interests:
Research Interests:
Research Interests:
Research Interests:
We propose an optimal choice for the bandwidth parameter used in the local polynomial estimation of the conditional distribution function. This choice is approximated by the bandwidth which minimizes the asymptotic weighted mean... more
We propose an optimal choice for the bandwidth parameter used in the local polynomial estimation of the conditional distribution function. This choice is approximated by the bandwidth which minimizes the asymptotic weighted mean integrated squared error (MISE). To cite this article: S. Ferrigno, G.R. Ducharme, C. R. Acad. Sci. Paris, Ser. I 346 (2008).
Research Interests:
ABSTRACT The asymptotic distribution of the circular median is derived for symmetric distributions on the circle. Its asymptotic relative efficiency with respect to the mean direction and to an stimator proposed by G. S. Watson... more
ABSTRACT The asymptotic distribution of the circular median is derived for symmetric distributions on the circle. Its asymptotic relative efficiency with respect to the mean direction and to an stimator proposed by G. S. Watson [Statistics on spheres (1983; Zbl 0646.62045)] is then examined. Special attention is given to the cases where the underlying distribution is von Mises and contaminated von Mises. It is seen that the circular median can perform more efficiently than both estimators in presence of outliers.
Research Interests:
A measure of multivariate correlation between two sets of vectors is considered when the underlying joint distribution is a member of the class of elliptical distributions. Its asymptotic distribution is derived under different situations... more
A measure of multivariate correlation between two sets of vectors is considered when the underlying joint distribution is a member of the class of elliptical distributions. Its asymptotic distribution is derived under different situations and these results are used to test hypotheses on vector correlation when the underlying joint distribution is non-normal.
Research Interests:
Research Interests:
Research Interests:
Biometrika (1995), 82, 2, pp. 437-8 Printed in Great Britain ... A note on smooth tests of goodness of fit for ... Departement de mathematiques et statistique, Universite de Montreal, CP 6128, Succ. "Centre-Ville", Montreal,... more
Biometrika (1995), 82, 2, pp. 437-8 Printed in Great Britain ... A note on smooth tests of goodness of fit for ... Departement de mathematiques et statistique, Universite de Montreal, CP 6128, Succ. "Centre-Ville", Montreal, Quebec, Canada, H3C 3J7
Research Interests:
Research Interests:
Research Interests:
Abstract. Directional data analysis usually assumes that the observations are recorded according to a coordinate system whose origin coincides with the center of their support. However, it may happen that the observer does not sit at this... more
Abstract. Directional data analysis usually assumes that the observations are recorded according to a coordinate system whose origin coincides with the center of their support. However, it may happen that the observer does not sit at this center and record the directions from another ...
Research Interests:
Research Interests:
Research Interests:
Since heart rate influences the values of many hemodynamic parameters, the authors have examined the effect of heart rate variations noted after cardiac transplantation on the hemodynamic criteria of rejection. They have undertaken serial... more
Since heart rate influences the values of many hemodynamic parameters, the authors have examined the effect of heart rate variations noted after cardiac transplantation on the hemodynamic criteria of rejection. They have undertaken serial studies at intrinsic rate and at a paced atrial rate of 150 beats/min in 11 resting homotransplanted dogs. The authors analysed cardiac index, stroke volume, peak velocity, acceleration, stroke power and stroke work. Results at intrinsic and paced atrial rates were compared during each of three phases of postoperative evolution: phase II (2 to 3 days after transplantation), phase III (4 to 5 days) and phase IV (6 to 7 days postoperatively). Phase II, being the period of maximum recovery, served as a baseline for comparing the cardiac parameters in phases III (the initial phase of rejection) and phase IV (24 hours before death). The authors found that the patterns of change were similar at intrinsic and paced atrial rates and concluded that heart rate variation does not significantly influence the hemodynamic criteria of rejection.
Research Interests:
Until recently intra-abdominal sepsis following surgery for perforated appendicitis has remained a frequent occurrence. In 1975 Fowler reported an incidence of one intra-abdominal abscess in 36 perforated appendicitis treated with a... more
Until recently intra-abdominal sepsis following surgery for perforated appendicitis has remained a frequent occurrence. In 1975 Fowler reported an incidence of one intra-abdominal abscess in 36 perforated appendicitis treated with a protocol consisting of saline irrigation at surgery, followed by intraperitoneal administration of cephaloridine q. 6 hours for 48 hours. The same antibiotic was then given i.v. for 72 hours. He concluded that this method was an effective way to minimize post-op intra-abdominal abscess, and that the intraperitoneal route was more effective, for this purpose, than systemic administration. More recently, with better knowledge of the role of anaerobes, improved results have also been obtained by adding clindamycin or metronidazole to the antibiotic regimen. A prospective study based on two groups of patients randomly assigned to two protocols was undertaken. The study group consisted of 45 children operated for perforated appendicitis with abscess or generalized peritonitis. These patients were treated with Fowler's protocol. Because it is less nephrotoxic and is active against the same organisms, cefazolin was used instead of cephaloridine. The control group consisted of 43 patients, also operated for perforated appendicitis, treated with saline irrigation at surgery, with either i.v. gentamycin or tobramycin and clindamycin or metronidazole for an average of 7.8 +/- 3.1 days. Both groups were fairly well matched in terms of age, sex, severity of disease, and bacterial flora. They showed an identical 6% post-op intra-abdominal sepsis rate. Days before afebrile, days under gastric suction, complications and average hospital stay were comparable in both groups. The wound infection rate was 20% in the study group and 9% in the control group.(ABSTRACT TRUNCATED AT 250 WORDS)
Research Interests:
Research Interests:
Research Interests:
In order to assess possible changes in myocardial relaxation occurring during the neonatal period, M-mode echocardiograms were recorded serially in 9 normal term infants and in another group of 10 one-month-old infants. The tracings were... more
In order to assess possible changes in myocardial relaxation occurring during the neonatal period, M-mode echocardiograms were recorded serially in 9 normal term infants and in another group of 10 one-month-old infants. The tracings were studied with an M-mode calculator. Although individual variations were greater in the data collected during the first 24 h, no significant difference was found in the indices of diastolic function of the left ventricle during the first 4 days of age. The following changes were observed between data recorded at 4 days and 1 month, respectively: normalized peak rate of left ventricle filling, 4.03 vs. 4.71 cm/s; diastolic peak velocity of early posterior motion of aortic root, 1.89 vs. 5.15 cm/s; peak velocity of left ventricle posterior wall motion in diastole, 3.31 vs. 3.50 cm/s; mitral valve EF slope, 59.05 vs. 84.92 mm/s; left ventricle isometric relaxation time, 43.88 vs. 28.50 ms. (1) greater individual variations are observed in indices of left ventricle diastolic function during the first day of life, and (2) significant increase in left ventricle compliance occurs during the first month of life. These changes should play a critical role in the clinical course of newborn with cardiopulmonary disease.
Research Interests:
Research Interests:
Sotalol is a beta-blocker with class III antiarrhythmic properties that has recently been used in children for the treatment of supraventricular and ventricular arrhythmias. However, little is known about its electrophysiologic effects on... more
Sotalol is a beta-blocker with class III antiarrhythmic properties that has recently been used in children for the treatment of supraventricular and ventricular arrhythmias. However, little is known about its electrophysiologic effects on the immature heart. Using intracardiac electrocardiographic recordings and stimulation techniques, 15 canine neonates (8-15 days) and 15 adult mongrel dogs were studied with cumulative doses of sotalol (0.5, 1, 2, and 4 mg/kg plus an additional dose of 8 mg/kg for neonates). Heart rate decreased significantly in the two groups, but more in adult dogs (-43% in adult dogs versus -25% in neonates, p less than 0.05). There was no significant change for QRS duration and His-Purkinje system conduction time interval. QT and atrioventricular nodal conduction time intervals increased in adult dogs and neonates. Sinus node recovery time increased significantly in the two groups, but more in adult dogs. Refractory periods of the atrioventricular (AV) node increased significantly in neonates. Atrial flutter was no longer inducible in 12 of 15 neonates after the 2 mg/kg dose. Atrial effective refractory period increased significantly more in neonates (96%, p less than 0.001) than in adult dogs (58%, p less than 0.001). Ventricular effective refractory periods increased significantly both in neonates (46%) and adult dogs (50%), in a similar way. In conclusion, sotalol has greater electrophysiologic effects on the immature heart at the atrial level when compared to the adult, and similar effects on the refractory period of AV node and ventricle.
Research Interests:
In this paper, we study the problem of estimating non-parametrically a quantile regression curve as it applies to computing reference values. We propose an automatic procedure that uses a symmetrized nearest-neighbor kernel estimator of... more
In this paper, we study the problem of estimating non-parametrically a quantile regression curve as it applies to computing reference values. We propose an automatic procedure that uses a symmetrized nearest-neighbor kernel estimator of conditional distributions. We also discuss ways of measuring the dispersion of quantile regression estimator. One is based on the asymptotic distribution of such quantiles, while the other relies on the bootstrap method. The results of a small simulation study show that the methods of the paper perform rather well even in a situation where a good parametric solution is available. As an example, we analyze a small part of a data set that was collected to establish reference values for blood velocity in different parts of the umbilical cord of human fetuse as they grow toward birth.
Research Interests:
The notion of efficient test for a Euclidean parameter in a semiparametric model was introduced by Stein [Proc. Third Berkeley Symp. Math. Statist. Probab. 1 (1956) 187-195]. Such tests are locally most powerful for a wide class of... more
The notion of efficient test for a Euclidean parameter in a semiparametric model was introduced by Stein [Proc. Third Berkeley Symp. Math. Statist. Probab. 1 (1956) 187-195]. Such tests are locally most powerful for a wide class of infinite-dimensional nuisance parameters. The first formal application of this notion to a suitably parametrized two-sample problem was provided by Hájek [Ann. Math. Statist. 33 (1962) 1124-1147]. However, this and subsequent solutions appear to be not well-suited for practical applications. This article aims to show that an adaptive two-sample test introduced recently by Janic-Wróblewska and Ledwina [Scand. J. Statist. 27 (2000) 281-297] is locally most powerful under a more realistic setting.