Search

INTRODUCTION: the interpretation of confidence intervals and hypothesis tests is usually directional, drawing a distinction between superiority and inferiority. On another note, numerous statistical tools are conditional to statistics that are actually variable such as the model matrix of a generalized linear model or sample size. The aim of this thesis is to propose methods for the assessment of statistical estimators that are consistent with the use that will be done of these tools. METHODS: the consequences of ignoring the directional interpretation in the assessment of statistical tools, especially for survival analysis, were evaluated from reviews of the literature and theoretical considerations. Assessment criteria Assessment criteria based on local averages were proposed to free the assessment of statistical tools from conditioning on variable statistics, such as sample size even when such tools are conditional. Underestimation and overestimation risks were differentiated in the assessment of coverage flaws of confidence intervals. Usual estimators were reassessed with these new criteria : binomial proportion and Poisson regression estimators. RESULTS: the directional interpretation of hypothesis tests about differences in survival curves can lead to a risk of concluding in the opposite direction to the actual difference approaching 50% in some cases. New assessment criteria, in particular, one-sided local average type I error rates, showed the better behavior of the likelihood ratio test and confidence intervals compared to Rao’s score and Wald’s methods for logistic and Poisson regressions. Strictly conservative methods are relevant in some very specific contexts ; otherwise, methods based on mid-P-values should be preferred. Confidence intervals aiming at an imbalance between risks of overestimation and underestimation in order to shorten their overall width lose directional interpretation properties. CONCLUSION: the directional interpretation of hypothesis tests and confidence interval estimators should be taken into account in their design and assessment., INTRODUCTION : l’interprétation des intervalles de confiance et des tests d’hypothèses est généralement directionnelle, c’est-à-dire, distinguant la supériorité de l’infériorité. Par ailleurs, de nombreux outils statistiques conditionnent sur des statistiques qui sont variables, tel que la matrice d’exposition d’un modèle linéaire généralisé ou la taille d’échantillon. L’objectif de cette thèse est de proposer des outils d’évaluation des estimateurs statistiques en adéquation avec l’usage de ces outils. MÉTHODES : les conséquences de l’absence de prise en compte de l’interprétation directionnelle dans l’évaluation des outils statistiques, notamment les tests d’hypothèses concernant les analyses de survie, ont été évaluées à partir de revues de la littérature et de considérations théoriques. Des critères de jugement, basés sur les moyennes locales ont été proposés afin de s’abstraire du conditionnement sur des statistiques variables, notamment la taille d’échantillon, lors de l’évaluation des outils statistiques. Les risques de sous-estimation et surestimation ont été distingués dans l’évaluation des défauts de couverture des intervalles de confiance. Des estimateurs classiques ont été réévalués à la lumière de ces nouveaux critères de jugement : estimateurs d’une proportion binomiale et estimateurs de régressions de Poisson. RÉSULTATS : l’interprétation directionnelle de tests d’hypothèses concernant les différences de courbes de survie peut conduire à un risque de conclure à une différence dans le sens opposé de la différence réelle, qui approche parfois 50 %. Les nouveaux critères de jugement, notamment les risques alpha unilatéraux moyens locaux, ont mis en évidence la supériorité du test du rapport de vraisemblance et des intervalles de confiance associés, par rapport aux méthodes du score de Rao et de Wald, pour les régressions logistiques et de Poisson. Les méthodes visant le strict conservatisme s’avèrent pertinentes dans certains contextes très spécifiques ; autrement les méthodes basées sur les mid-P-valeurs devraient leur être préférées. Les intervalles de confiance visant à un déséquilibre des risques de surestimation et de sous-estimation afin d’en rétrécir la largeur totale perdent des propriétés d’interprétation directionnelle.CONCLUSION : l’interprétation directionnelle qui sera faite des tests d’hypothèses et estimateurs d’intervalles de confiance devrait être prise en compte lors de leur conception et évaluation.

This paper examines the cross-linguistic phenomenon of locative case restricted to a closed class of items (L-nouns). Starting with Latin, I suggest that the restriction is semantic in nature: L-nouns denote in the spatial domain and hence can be used as locatives without further material. I show how the independently motivated hypothesis that directional PPs consist of two layers, Path and Place, explains the directional uses of L-nouns and the cases that are assigned then, and locate the source of the locative case itself in p0, for which I then provide a clear semantic contribution: a type-shift from the domain of loci to the object domain. I then examine cross-linguistic restrictions on the use of locative case and show that the patterns observed can be accounted for on the same assumptions., Proceedings of Sinn und Bedeutung, Vol 23 No 2 (2019): Proceedings of Sinn und Bedeutung 23

International audience; Nêlêmwa is an Aspect-Mood oriented language; verbs are unmarked for tense, time reference is expressed by chronology and time adverbs. Aspect hinges on three notions: events (in the aorist), states, and processes. One focus is the contrast between the perfect and the aorist. Bare aorist verb forms refer to events or to sequences of events with no reference to their internal phases. The perfect expresses internal relations between processes and clauses (anteriority, backgrounding, causal relations); it refers to transitional processes that have reached or not their final instant, expressing changes of states and resulting states. In future reference frames, the perfect expresses imminent change of states, or imminent completion of a process, and the speaker's certainty about their projected occurrence.

11 pages et 4 plans; Directional drilling tool (13, 14) comprising a cutting structure formed of blades on which there are cutting edges (13) and a guard (11) that stabilizes the tool against the wall of the well (3), characterized in that the guard comprises a plurality of passive pivoting blades (16) each of which can pivot about a pivot pin (18) inserted in the guard (12) near its boundary with the cutting structure under the action of thrusting and/or damping members, so that when the pivoting blades (16) are in their nominal position, the guard (12) of the tool (1, 14) exhibits a substantially cylindrical shape

International audience; Bi-MnBi composite has been directionally solidified to investigate the influence of a strong magnetic field on its microstructure, growth characteristics and magnetic properties. Experimental results show that MnBi eutectic fiber is aligned, regularized and coarsened during directional solidification under an axial strong magnetic field. In particular, the alignment of irregular eutectic fiber under a magnetic field may offer a new method to prepare regular fiber composite and extend the range of uses of such composite. It has also been found that the strong magnetic field has modified the growth mode of the MnBi fiber and enhanced the transition from non-faceted to faceted. The distribution of the solute Mn in front of the MnBi fiber at the liquid solid interface has been measured and the result shows that the magnetic field has enhanced the enrichment of the solute Mn in front of the MnBi fiber. Moreover, magnetic measurement reveals a pronounced promotion of the magnetic anisotropy and the remanence (B(r)) under the magnetic field. Based on these experimental results, a simple theory of solute enrichment in front of a ferromagnetic fiber and fiber coarsening under a strong magnetic field has been proposed. (C) 2011 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved

The invention concerns a method which consists in: introducing into a pre-drilled portion of a well (16) a drilling system comprising a drilling bit (11) assembled with tubular lining elements (15), means for measuring the inclination and azimuth direction of said pre-drilled part of the well, means for rotary friction (26, 27) against the wall of said pre-drilled part and controlled means capable of modifying the friction coefficient of said friction means against the wall; measuring the azimuth direction of the pre-drilled part of the well (16); and varying the friction coefficient of said friction means (26, 27) sufficiently to influence the azimuth direction of the next well portion to be drilled.

In recent years, we proposed a number of new and very effective features for automatic writer identification and verification. They are probability distribution functions (PDFs) extracted from the handwriting images and characterize writer individuality independently of the textual content of the written samples. In this paper, we perform an extensive analysis of feature combinations. In our fusion scheme, the final unique distance between two handwritten samples is computed as the average of the distances due to the individual features participating in the combination. Obtained on a large dataset containing 900 writers, our results show that fusing multiple features (directional, grapheme, run-length PDFs) yields increased writer identification and verification performance.

Summary The importance of wellbore deviation is well recognized by the drilling industry. An analysis of a drilling system's directional behavior must include the directional characteristics of the drill bit. This paper presents a comprehensive analysis of the directional behavior of polycrystalline diamond compact (PDC) bits, including the effect of bit profile, gauge cutters, and gauge length. Numerical simulations as well as laboratory tests have been carried out to better understand the mechanisms of PDC bit deviation and to evaluate the most important parameters affecting the directional behavior of PDC bits. The analysis presented in this paper shows that each part of the PDC bit (profile and active and passive gauges) plays a major role in its walking tendency and steerability. A quantitative evaluation of how these factors contribute to well trajectory (inclination and azimuth) is given. For the first time, a full-scale directional-drilling bench was built to measure the walking tendency and steerability of PDC bits. The results obtained demonstrate that the bit profile, gauge cutters, and gauge length have a significant effect. A 3D theoretical rockbit interaction model was developed to reproduce the drilling test results.