La mesure de la phase phis violant la symétrie CP dans les désintégrations B_s -> J/psi phi est une des analyses phares de l'expérience LHCb. Le Modèle Standard prédit cette observable avec une faible incertitude théorique. La présence de Nouvelle Physique notamment dans les diagrammes en boucle de l'oscillation B_s - B_sbar peut significativement modifier la mesure de cette phase. Durant cette thèse, nous avons participé à différents aspects de l'analyse des désintégrations B_s -> J/psi phi. Nous avons proposé une sélection multi-variable, basée sur un algorithme d'arbres de décision boostés. Cette nouvelle sélection augmente le nombre de signal B_s -> J/psi phi de 19% mais nécessite une correction temporelle. Une étude portant sur l'origine des distorsions angulaires caractéristiques des désintégrations B_s -> J/psi phi a aussi été menée. Il s'avère que la principale source de distorsion est la géométrie du détecteur LHCb. Nous avons aussi contribué à un autre élément clef de cette analyse : l'étiquetage de la saveur. Afin de connaître la saveur initiale des mésons B, un algorithme d'étiquetage a été développé. Il exploite les traces chargées provenant du hadron B de côté opposé. Nous avons participé à son optimisation ainsi qu'à son étalonnage en étudiant les désintégrations B_d -> J/psi K*. En utilisant l'intégralité des données enregistrées en 2011, soit 1 fb^{-1}, après optimisation et calibration de cet algorithme, la puissance d'étiquetage dans le canal B_s -> J/psi phi est : (2.29 +- 0.07 +- 0.26) %. La valeur de phis mesurée est alors : phis = -0.001 +- 0.101 +- 0.027 rad dans les désintégrations B_s -> J/psi phi. / The measurement of the CP violating phis phase in the B_s -> J/psi phi decays is one of the main goals of the LHCb experiment. The predicted value of this observable by the Standard Model has a small theoretical uncertainty. New Physics could significantly modify this measurement, with the presence of new particles in the B_s-B_sbar oscillation's box diagrams. During this thesis, we participate to many aspects of B_s -> J/psi phi analysis. We proposed a multivariate selection, based on a boosted decision trees algorithm. This new selection adds 19% signal B_s -> J/psi phi but it needs a time correction. We studied the main sources of angular distorsions of B_s -> J/psi phi decays. We found that the detector's geometry has the most important impact on the angular distributions. We participated to another key element of the phis analysis : the flavour tagging. In order to determine the initial B flavour, a tagging algorithm has been developped. It exploits charged tracks originating from the b-hadron opposite to the signal B-meson. We participated to its optimization and calibration, by studing the B_d -> J/psi K* channel. With 1 fb^{-1} corresponding to the whole 2011 data, after the optimization and calibration of this algorithm, the tagging power is measured in the B_s -> J/psi phi channel : (2.29 +- 0.07 +- 0.26) %. The phis measurement in the B_s -> J/psi phi channel is : phis = -0.001 +- 0.101 +- 0.027 rad. The combined analysis of B_s -> J/psi phi and B_s -> J/psi pi^+ pi^- decays gives the most precise measurement of phis = -0.002 +- 0.083 +- 0.027 rad. This measurement is compatible with the Standard Model, but there is still room for New Physics.

Sei $K$ eine endliche Erweiterung von $QQ_p $ und sei $R$ der Robba-Ring mit Koeffizienten in $K$ sein, die mit einem absoluten Frobenius-Lift $phi$ ausgestattet sind. Sei $F$ der Fixköper von $K$ unter $phi $ und sei $G$ eine verbundene reduktive Gruppe über $F$. Diese Arbeit untersucht $G$-$ (phi,nabla)$-Module über $R$, nämlich $(phi,nabla)$-Module über $R$ mit einer zusätzlicher $G$-Struktur. In Kapitel 3 konstruieren wir einen gefilterten Faserfunktor aus der Darstellungskategorie von $G$ auf endlich-dimensionalen $F$-Vektorräumenbis zur Kategorie von $QQ$-gefilterten Modulen über $R$, und beweisen, dass dieser Funktor spaltbar ist. In Kapitel 4 beweisen wir eine $G$-Version des $p$-adischen lokalen Monodromie-Satzes. In Kapitel 5 beweisen wir eine $G$-Version des logarithmischen lokalen Monodromie-Satzes unter bestimmten Annahmen. Als Anwendung fügen wir jedem $G$-$(phi,nabla)$-Modul eine Weil-Deligne-Darstellung der Weil-Gruppe $W_{kk((t))} $ in $G(K^{nr})$ an, wobei $kk$ der Restklassenkörper von $K$, und $K^{nr}$ die maximal unverzweigte Erweiterung von $K$ ist. / Let $K$ be a finite extension of $QQ_p$ and let $R$ be the Robba ring with coefficients in $K$, equipped with an absolute Frobenius lift $phi$. Let $F$ be the fixed field of $K$ under $phi$ and let $G$ be a connected reductive group over $F$. This thesis investigates $G$-$(phi,nabla)$-modules over $R$, namely $(phi,nabla)$-modules over $R$ with an additional $G$-structure. In Chapter 3, we construct a filtered fiber functor from the category of representations of $G$ on finite-dimensional $F$-vector spaces to the category of $QQ$-filtered modules over $R$, and prove that this functor is splittable. In Chapter 4, we prove a $G$-version of the $p$-adic local monodromy theorem. In Chapter 5, we prove a $G$-version of the logarithmic $p$-adic local monodromy theorem under certain assumptions. As an application, we attach to each $G$-$(phi,nabla)$-module a Weil-Deligne representation of the Weil group $W_{kk((t))}$ into $G(K^{nr})$, where $kk$ is the residue field of $K$, and $K^{nr}$ is the maximal unramified extension of $K$.

In aspiration to achieve naturalistic character animation, some difficulty does show up, both in 3D-CGI, 2D-CGI as in traditional handmade animation. The reason why difficulty appears in the task of achieving naturalism is for instance because of the simulated natural laws in the digital world which cannot affect the characters cinematic movement. The result of illustrating the natural law manually by hand can affect the illusion of the movement as not naturalistic or inconsistent, which can result in that the spatial adaptation to space is not perceived as naturalistic. It is most often up to the animator to make a believable illusion of the natural laws. This study examines a deeper understanding of the human way to behave and to adapt to the physical reality and through that, apply this movement on character animation. The purpose of a deeper understanding of movement's phenomena is to find a clearer method dedicated to naturalistic movement that can be adapted to all forms of character animation, regardless of the character anatomy. Unfortunately, it would not be possible to study all aspects of the physically adaptive behaviors as to the size of the topic. This study will therefore orbit the human locomotion, also known as walking. Focus will therefore be on the bottom part of the body, from the pelvis down to the ankles. There are scientists that do argue about a connection between the golden ratio and the human locomotion. This study is an attempt to apply the golden ratio on 3D-characters. This study has been devoted to a different approach in the character animation, where its starting point will be mathematical equations. For this study, it has meant an application of the golden section in a temporary form, that is in motion. The result is therefore a mathematically based method dedicated to imitating naturalistic movement in terms of time, that is to say human locomotion. The method has been proved to be useful even in the imitation of characters movements that do not possess human proportions. / I en strävan efter naturalistisk karaktäranimation uppstår vissa hinder på vägen i 3D-CGI som 2D-CGI animation. Svårigheter i att uppnå naturalism beror bland annat på att det inte finns några simulerade naturlagar i den digitala världen som kan påverka karaktärens kinematiska rörelser. Resultatet av att behöva illustrera naturlagar och liknande externa påverkningar manuellt för hand kan orsaka en rumslig känsla om onaturlighet eller inkonsekvens. Vilket kan resultera i att den rumsliga anpassningen till rummet inte upplevs som naturalistisk. Ofta är det upp till animatören att skapa en trovärdig illusion om naturlagarna. Målet med denna studie är att utforska och ge animatören en djupare förståelse för människans sätt att förhålla och anpassa sig till den fysiska verkligheten. Syftet med en djupare förståelse för rörelsers företeelser är att finna en tydligare metod dedikerad åt naturalistisk rörelse som går att anpassa för alla former av karaktärsanimation, oberoende av karaktärens anatomi. Tyvärr skulle det inte vara möjligt att täcka alla aspekter av det fysiskt adaptiva beteendet på grund av ämnets storlek. Denna studie har därför avgränsats till mänsklig lokomotion, även känt som vandring. Fokus riktas därför mot bäckenet och knäleden ner till anklarna. Det finns forskning som argumenterar för en koppling mellan det gyllene snittet och mänskliga rörelser. Denna studie är ett försök att tillämpa lokomotion uträknat ifrån gyllene snittet på en digital tredimensionell karaktär. Denna studie har ägnats ett annorlunda angreppsätt i karaktärsanimationen där dess utgångspunkt består av matematiska ekvationer. Det har för studien inneburit ett applicerande av gyllene snittet i en temporär form, det vill säga i rörelse. Resultatet är därför en matematiskt baserad metod dedikerad för att efterlikna naturalistisk rörelse i form av gång, alltså mänsklig lokomotion. Metoden har visat sig användbar även vid efterliknandet av karaktärers rörelser som inte besitter mänskliga proportioner.

Cette thèse est constituée de deux parties indépendantes, étudiant deux aspects de la théorie des (φ,Γ)-modules en caractéristique p. La première partie porte sur l'étude de la réduction modulo p des représentations cristallines irréductibles de dimension deux. Nous donnons, pour des poids k ≤ p², un calcul explicite de la réduction de V(k,a) pour a dans un disque fermé centré en zéro, généralisant ainsi des résultats déjà connus pour k ≤ 2p. En particulier, nous calculons le plus grand rayon possible pour ce disque, et montrons que dans certains cas, la réduction qui est constante à l'intérieur du disque change sur son bord. Dans la seconde partie, nous nous intéressons aux représentations d'un sous-groupe de Borel de GL[indice]2(Q[indice]p) sur un corps de caractéristique p, et en particulier à celles qui sont lisses, irréductibles et admettent un caractère central. Une méthode pour construire de telles représentations à partir de (φ,Γ)-modules irréductibles a été décrite par Colmez dans sa construction de la correspondance de Langlands p-adique. Après avoir donné un cadre un peu plus général dans lequel la construction de Colmez fonctionne encore, nous classifions les représentations irréductibles du Borel, prouvant que la construction précédente permet d'obtenir toutes les représentations de dimension infinie. Lorsque le corps des coefficients est fini, ou algébriquement clos, nous disposons d'une interprétation galoisienne des (φ,Γ)-modules irréductibles, et la classification précédente permet alors d'obtenir une correspondance entre ces représentations du Borel et des représentations galoisiennes modulaires. / This thesis is made of two independent parts, dealing with two different aspects of characteristic p (φ,Γ)-modules. In the first part we study the reduction modulo p of -2-dimensional irreducible crystalline representations. For weights k ≤ p2, we give an explicit description of the reduction V(k,a) for a belonging to a closed disk centered at zero, generalizing results already known for k ≤ 2p. We explicitely compute the biggest possible radius for this disk, and prove that in some cases, the reduction which is constant on the interior of the disk is different for a belonging to the border of the disk. In the second part, we study the smooth, irreducible representations of a Borel subgroup of GL[indice]2(Q[indice]p) over a field of characteristic p and admitting a central character. One way of constructing such representations from irreducible (φ,Γ)-modules was described by Colmez in his construction of the p-adic Langlands correspondence. After giving a more general framework for Colmez's construction, we classify the irreducible representations of the Borel subgroup, proving that the previous construction already gives all the infinite dimensional representations. When the coefficient field is finite, Fontaine's equivalence combined with the previous classification gives a correspondence between these representations of a Borel subgroup of GL[indice]2(Q[indice]p) and modular galois representations.

Les glutathion transférases (GSTs) constituent une famille multigénique d’enzymes présentes dans les trois domaines du vivant. Cette présence ubiquitaire souligne l’origine sans doute très ancienne de ces enzymes ainsi que des fonctions fondamentales conservées au cours de l’évolution. Ces enzymes sont impliquées notamment dans la détoxication cellulaire de molécules toxiques et dans le métabolisme secondaire. Les analyses phylogénétiques regroupent les GSTs des organismes photosynthétiques au sein de quatorze classes qui peuvent être séparées en deux grands groupes selon le résidu catalytique : les GSTs à sérine catalytique qui possèdent des activités de conjugaison du glutathion (GSH) et/ou peroxydase tandis que les GSTs à cystéine catalytique présentent des activités thioltransférase, déshydroascorbate réductase et de déglutathionylation. Les GSTs à sérine catalytique de la classe Phi (GSTF) sont présentes chez les organismes photosynthétiques et certains basidiomycètes. Chez les plantes, cette classe comprend un nombre de gènes plus important que les autres classes de GSTs. Ceux-ci sont parmi les plus régulés en réponse à divers stress et ils ont été fortement étudiés chez les plantes céréalières en raison de l’activité de détoxication des herbicides des protéines correspondantes. Pourtant, à quelques exceptions près, les rôles physiologiques des GSTFs restent inconnus et la redondance d’isoformes dans cette classe reste incomprise. Par des approches moléculaires, biochimiques et structurales, l’analyse structure-fonction des huit GSTFs de l’arbre modèle Populus trichocarpa a été réalisée au cours de cette thèse. L’analyse phylogénétique des GSTFs chez les organismes photosynthétiques a montré que cette classe est apparue au moment de l’apparition terrestre des végétaux et que différents groupes pouvaient être identifiés avec des motifs catalytiques distincts. L’analyse transcriptionnelle a montré que les gènes relatifs aux GSTFs de peuplier sont principalement exprimés dans les fleurs femelles, les pétioles et les fruits. Certains aspects du mécanisme réactionnel ont été caractérisés en déterminant notamment les paramètres cinétiques et d’interaction des huit GSTFs et de plusieurs variants mutés pour des résidus clés vis-à-vis de substrats modèles. Les structures de cinq des huit GSTFs ont été résolues et ces protéines dimériques adoptent un repliement GST canonique et des spécificités structurales au niveau du site actif ont pu être observées. De plus, au regard de la capacité des orthologues des GSTFs à lier des hormones et des flavonoïdes ainsi que de l’expression récurrente des GSTFs de peuplier dans les fruits et les fleurs femelles, deux organes riches en ces molécules, il peut être supposé qu’elles ont aussi des propriétés de type ligandine. Des résultats préliminaires ont également été obtenus pour la recherche de substrats physiologiques à partir de métabolites extraits de différents organes de peuplier. A terme, l’identification de ces substrats permettra de déterminer le mode d’action (catalytique vs ligandine) de chaque enzyme et d’identifier clairement les fonctions in planta de ces enzymes / Glutathione transferases (GSTs) belong to a multigenic family whose presence in most eukaryotes, prokaryotes and archaea reflects their widespread nature and very likely important functions. These enzymes represent a major group of enzymes involved in xenobiotic detoxification and secondary metabolism. From the most recent genomic and phylogenetic analyses, the GST family is subdivided into 14 classes that can be separated into two main groups based on the catalytic residue which is either a serine (Ser-GST) or a cysteine (Cys-GST). Ser-GSTs usually catalyze glutathione (GSH) conjugation and/or peroxide reduction. On the other hand, Cys-GSTs cannot perform GSH-conjugation reactions but instead catalyze thiol-transferase, dehydroascorbate reductase and deglutathionylation reactions. Ser-GSTs from the Phi class (GSTF) are present in photosynthetic organisms and some basidiomycetes. This class is composed of a large number of genes compared to other GST classes which are amongst the most stress-inducible. The corresponding proteins have been extensively studied in crops with regard to their detoxification activities toward herbicides. However, with a few exceptions, very little is known about their roles in planta and it is not well understood why this class has expanded. By combining molecular, cellular, biochemical and structural approaches, the eight isoforms from the model tree Populus trichocarpa have been characterized during this PhD project. Phylogenetic analysis of GSTFs in the green lineage shows that the apparition of this class is concomitant with the appearance of terrestrial plants and that different groups can be distinguished based on the active site signature. RT-PCR analysis of the eight isoforms of GSTFs showed that transcripts mostly accumulate in female flowers, petioles and fruits. Some aspects of the reaction mechanism have been characterized by determining kinetic parameters of the eight poplar GSTFs and of several mutated variants for key residues towards model substrates. The structures of five GSTFs have been solved and these dimeric proteins display a typical GST fold but specificities have been observed at the catalytic site level. Moreover, considering the demonstrated capacity of GSTF orthologs to bind hormones, anthocyanins or flavonoids, and the consistent high expression of poplar GSTFs in female flowers and fruits, two organs rich in these molecules, we speculate that they may also possess ligandin properties. Preliminary results have been obtained regarding the nature of the substrates in various poplar organs by analyzing protein thermostability in the presence of putative ligands. In order to assess whether a functional redundancy between poplar Phi GSTs exists and to identify their mode of action (catalytic vs ligandin functions), we started to isolate and identify physiological substrates

No description available.

International Telemetering Conference Proceedings / October 28-31, 1996 / Town and Country Hotel and Convention Center, San Diego, California / The use of high bit rates in the missile testing environment requires that the receiving telemetry system(s) have the correct signal margin for no PCM bit errors. This requirement plus the fact that the use of “redundant systems” are no longer considered optimum support scenarios has made it necessary to select the minimum number of tracking sites that will gather the data with the required signal margin. A very basic link analysis can be made by using the maximum and minimum gain values from the transmitting antenna pattern. Another way of evaluating the transmitting antenna gain is to base the gain on the highest percentile appearance of the highest gain value. This paper discusses the mathematical analysis the WSMR Telemetry Branch uses to determine the signal margin resulting from a radiating source along a nominal trajectory. The mathematical analysis calculates the missile aspect angles (Theta, Phi, and Alpha) to the telemetry tracking system that yields the transmitting antenna gain. The gain is obtained from the Antenna Radiation Distribution Table (ARDT) that is stored in a computer file. An entire trajectory can be evaluated for signal margin before an actual flight. The expected signal strength level can be compared to the actual signal strength level from the flight. This information can be used to evaluate any plume effects.

This thesis investigates the patterning of the Algonquian “central agreement”, i.e. the primary person-number agreement marking, from a diachronic and comparative perspective. The central agreement patterns differ in the two orders: in the conjunct it is fusional and often portmanteau while in the independent it is discontinuous and non-portmanteau. In addition to these differences, there are also some commonalities, such as a pattern in which 1p consistently outranks 2p in both orders. This thesis shows that the differences between the two orders can be taken to reflect variation in the features of the syntactic probe and different morphological spell-out rules, while the shared properties follow from the underlying structure of φ-features. In particular, it is proposed that an additional person feature under the [plural] node causes first person plural to be privileged over second person plural in the competition among vocabulary items in post-syntactic spellout. / October 2016

This study compared three methods of selecting the best discriminating test items and the resultant test reliability of mastery/nonmastery classifications. These three methods were (a) the agreement approach, (b) the phi coefficient approach, and (c) the random selection approach. Test responses from 1,836 students on a 50-item physical science test were used, from which 90 distinct data sets were generated for analysis. These 90 data sets contained 10 replications of the combination of three different sample sizes (75, 150, and 300) and three different numbers of test items (15, 25, and 35). The results of this study indicated that the agreement approach was an appropriate method to be used for selecting criterion-referenced test items at the classroom level, while the phi coefficient approach was an appropriate method to be used at the district and/or state levels. The random selection method did not have similar characteristics in selecting test items and produced the lowest reliabilities, when compared with the agreement and the phi coefficient approaches.

Modern applications generate a massive amount of data that is challenging to process or analyse. Graph algorithms have emerged as a solution for the analysis of this data because they can represent the entities participating in the generation of large scale datasets in terms of vertices and their relationships in terms of edges. Graph analysis algorithms are used for finding patterns within these relationships, aiming to extract information to be further analysed. The breadth-first search (BFS) is one of the main graph search algorithms used for graph analysis and its optimisation has been widely researched using different parallel computers. However, the BFS parallelisation has been shown to be chal- lenging because of its inherent characteristics, including irregular memory access patterns, data dependencies and workload imbalance, that limit its scalability. This thesis investigates the optimisation of the BFS on the Xeon Phi, which is a modern parallel architecture provided with an advanced vector processor using a self-created development framework integrated with the Graph 500 benchmark. As a result, optimised parallel versions of two high-level algorithms for BFS were created using vectorisation, starting with the conventional top-down BFS algorithm and, building on this, leading to the hybrid BFS algorithm. The best implementations resulted in speedups of 1.37x and 1.33x, for a one million vertices graph, compared to the state-of-the-art, respectively. The hybrid BFS algorithm can be further used by other graph analysis algorithms and the lessons learned from vectorisation can be applied to other algorithms targeting the existing and future models of the Xeon Phi and other advanced vector architectures.