A pesquisa tem como foco a forma do livro na cultura da convergência, na sua relação entre design e edição, partindo do conceito de dispositivo de Charaudeau (2006a e 2006b) e Mouillaud (2002). Dada a amplitude da questão, buscou-se um objeto empírico que permitisse relacionar informações de caráter histórico, dados documentais e conceitos construídos com base no referencial teórico, ancorando a problematização proposta. Para tanto, selecionou-se o romance Memórias Póstumas de Brás Cubas, de Machado de Assis, editado pela primeira vez em livro em 1881. O objetivo geral foi identificar paratextos editoriais ligados à forma em oito edições publicadas nos séculos XIX, XX e XXI, impressas e digitais, a fim de analisar e discutir como esses articulam-se à noção de livro enquanto dispositivo, problematizando tais relações a partir das lógicas de funcionamento de sua cadeia e mercado, dos novos modelos de acesso às edições e do lugar ocupado pelo livro na cultura da convergência. Como objetivos específicos, buscouse (1) avaliar o papel do design no estabelecimento de paratextos editoriais, definindo níveis de atuação da atividade na conformação do livro enquanto dispositivo; (2) identificar e analisar os paratextos editoriais ligados à forma e ao entorno do livro em diferentes edições, impressas e digitais do título, tensionando o modo como as edições configuram materialmente orientações e tendências na cadeia do livro; e (3) problematizar a noção de livro na cultura da convergência, tendo em vista o processo de remediação no sistema de mídia. A análise das edições delineou-se em duas etapas. A primeira teve em vista os paratextos ligados à forma. Já na segunda etapa, focou-se na identificação de tendências nos paratextos ligados ao entorno do livro. Constatou-se que as edições analisadas variam muito em relação à forma, indicando que os produtos foram criados para leituras e leitores diversos. As edições do século XIX são objetos que integram o texto verbal e a retórica tipográfica, valorizando o espaço gráfico como elemento importante para a narrativa. Nas edições digitais do século XXI, há outras propostas, com novos atrativos, incluindo desde ferramentas de busca dentro do texto até a possibilidade de cruzamento do texto verbal na formação de epitextos também verbais. Entre imediação e hipermediação, a materialidade do livro hoje é resultado de um processo contínuo de remediação do livro impresso no digital e do digital no impresso. A materialidade é influenciada também pela noção de espaço nos produtos digitais. Esse entorno – que no livro impresso está claramente fora do produto – no livro em tela passa a ter seus limites menos claros, permeando o artefato digital. Notou-se, nas três edições digitais analisadas, que bibliotecas, livrarias e lojas virtuais geram epitextos cada vez mais aderentes aos produtos editoriais. O livro digital não tem ainda uma identidade própria, mas delineia uma tendência importante neste início de século XXI: o tratamento do livro como objeto de dados predominantemente verbais. Embora essa seja uma vertente importante do livro atual, ela não é a única. Convivem na cultura da convergência muitas formas, em papel e em tela, que adotam estruturas bastante similares entre si, cercadas de inúmeros outros modos de narrar essa história, em meio a leitores cada vez mais expostos e acostumados a narrativas transmídias. / The research focuses on the form of the book in the convergence culture in its relationship between design and editing, based on the concept of dispositif by Charaudeau (2006a and 2006b) and Mouillaud (2002). Given the breadth of the issue, it sought an empirical object that allows relating historical background information, documentary evidence and concepts built on the theoretical framework, anchoring the problematic proposal. To accomplish this, we selected the novel The Posthumous Memoirs of Bras Cubas, by Machado de Assis, published for the first time as a book in 1881. The overall goal was to identify editorial paratexts connected to the form of the book in eight editions, published in the 19th, 20th and 21st centuries, print and digital, in order to analyze and discuss how they articulate the notion of book as dispositif in the convergence culture. As specific objectives, we sought to (1) evaluate the role of design in establishing editorial paratexts defining levels of action in shaping the book as a dispositif; (2) identify and analyze the editorial paratexts on different book editions, print and digital, discussing how they materially configure trends in the book chain; and (3) discuss the notion of book on the convergence culture, considering the remediation process in the media system. The analysis of the books was outlined in two stages. The first one took into consideration the paratexts attached to the form. In the second phase, the focus was on identifying paratexts connected to the surroundings of the books. As a result, we identified that the editions vary widely in regard to form, indicating that the products were created to several different readings and readers. The books edited in the 19th century are objects that integrate the verbal text and the rhetoric of typography, highlighting the graphic space as an important element to the narrative. In the digital editions of the 21st century, there are other strategies, with new tools, ranging from search engines to the possibility of crossing verbal texts, what generates new verbal epitexts. Between immediacy and hypermediacy, the materiality of the book today is the result of a continuous process of remediation of the printed book into the digital book and vice-versa. In digital products, materiality is also influenced by the notion of space. The epitext environment, in the printed book, is clearly outside the product. In the digital book, on the other hand, the boundaries between what is inside and outside the book is not as clear. It was noticed that in the three digital editions analyzed, libraries, bookstores and online retailers generate epitexts that are adhered to the editorial products. The digital book does not have its own identity yet, but we see an important trend in the beginning of 21st century: the book treated as a predominantly verbal data object. While this is an important aspect of the current idea of book, it is not the only one. Books exist in the convergence culture in many ways, on paper and on digital screen. They have very similar structures and are surrounded by numerous other ways of telling Bras Cubas’ story, accessed by readers increasingly exposed and accustomed to transmedia narratives.

Dans cette thèse, nous étudions des stratégies d’allocation séquentielle de ressources. Le modèle statistique adopté dans ce cadre est celui du bandit stochastique à plusieurs bras. Dans ce modèle, lorsqu’un agent tire un bras du bandit, il reçoit pour récompense une réalisation d’une distribution de probabilité associée au bras. Nous nous intéressons à deux problèmes de bandit différents : la maximisation de la somme des récompenses et l’identification des meilleurs bras (où l’agent cherche à identifier le ou les bras conduisant à la meilleure récompense moyenne, sans subir de perte lorsqu’il tire un «mauvais» bras). Nous nous attachons à proposer pour ces deux objectifs des stratégies de tirage des bras, aussi appelées algorithmes de bandit, que l’on peut qualifier d’optimales. La maximisation des récompenses est équivalente à la minimisation d’une quantité appelée regret. Grâce à une borne inférieure asymptotique sur le regret d’une stratégie uniformément efficace établie par Lai et Robbins, on peut définir la notion d’algorithme asymptotiquement optimal comme un algorithme dont le regret atteint cette borne inférieure. Dans cette thèse, nous proposons pour deux algorithmes d’inspiration bayésienne, Bayes-UCB et Thompson Sampling, une analyse à temps fini dans le cadre des modèles de bandit à récompenses binaires, c’est-à-dire une majoration non asymptotique de leur regret. Cette majoration permetd’établir l’optimalité asymptotique des deux algorithmes. Dans le cadre de l’identification des meilleurs bras, on peut chercher à déterminer le nombre total d’échantillons des bras nécessaires pour identifier, avec forte probabilité, le ou les meilleurs bras, sans la contrainte de maximiser la somme des observations. Nous définissons deux termes de complexité pour l’identification des meilleurs bras dans deux cadres considérés dans la littérature, qui correspondent à un budget fixé ou à un niveau de confiance fixé. Nous proposons de nouvelles bornes inférieures sur ces complexités, et nous analysons de nouveaux algorithmes, dont certains atteignent les bornes inférieures dans des cas particuliers de modèles de bandit à deux bras, et peuvent donc être qualifiés d’optimaux. / In this thesis, we study strategies for sequential resource allocation, under the so-called stochastic multi-armed bandit model. In this model, when an agent draws an arm, he receives as a reward a realization from a probability distribution associated to the arm. In this document, we consider two different bandit problems. In the reward maximization objective, the agent aims at maximizing the sum of rewards obtained during his interaction with the bandit, whereas in the best arm identification objective, his goal is to find the set of m best arms (i.e. arms with highest mean reward), without suffering a loss when drawing ‘bad’ arms. For these two objectives, we propose strategies, also called bandit algorithms, that are optimal (or close to optimal), in a sense precised below. Maximizing the sum of rewards is equivalent to minimizing a quantity called regret. Thanks to an asymptotic lower bound on the regret of any uniformly efficient algorithm given by Lai and Robbins, one can define asymptotically optimal algorithms as algorithms whose regret reaches this lower bound. In this thesis, we propose, for two Bayesian algorithms, Bayes-UCB and Thompson Sampling, a finite-time analysis, that is a non-asymptotic upper bound on their regret, in the particular case of bandits with binary rewards. This upper bound allows to establish the asymptotic optimality of both algorithms. In the best arm identification framework, a possible goal is to determine the number of samples of the armsneeded to identify, with high probability, the set of m best arms. We define a notion of complexity for best arm identification in two different settings considered in the literature: the fixed-budget and fixed-confidence settings. We provide new lower bounds on these complexity terms and we analyse new algorithms, some of which reach the lower bound in particular cases of two-armed bandit models and are therefore optimal

L'une des stratégies mise en œuvre pour améliorer la prise en charge thérapeutique des patients concerne le développement d'outils diagnostiques sensibles et spécifiques. Les aptamères sont des oligonucléotides artificiels obtenus par SELEX avec une très haute affinité ainsi qu'une excellente spécificité pour leurs cibles. L'immobilisation de ces motifs de reconnaissance moléculaire à la surface de nanomatériaux tels que des nanoparticules d'or (AuNPs), dont les propriétés optiques et électroniques sont uniques, permet d'amplifier le signal généré par l'interaction du ligand avec sa cible. Deux systèmes de biosensing ont été élaboré en fonctionnalisant des AuNPs avec des aptamères, l'un dirigé contre la thrombine et le second dirigé contre une marque épigénétique portée par une protéine histone. La réduction des sels d'or aurique précurseurs a été réalisée en présence de PEG4 et a conduit à l'obtention d'une population homodisperse de AuNPs sphériques d'un diamètre moyen de 14 nm et présentant une isotropie de taille et de forme. Ces AuNPs ont ensuite été fonctionnalisées par des bras espaceurs de longueur variable constitués d'unités tétraéthylène glycol successives reliées entre elles par des ponts éthers ou triazoles. L'acide lipoique a été utilisé comme motif d'ancrage à la surface des AuNPs via une liaison covalente Au-S et a été couplé aux différents bras espaceurs via une réaction de Steglich. Les linkers étaient porteurs d'un groupement terminal azoture afin de réaliser le couplage par chimie-click avec les aptamères. La stratégie de détection de la thrombine utilisait les propriétés de quenching de fluorescence des AuNPs alors que la détection de l'histone était colorimétrique et mettait à profit l'effet de résonance plasmonique de surface des nanoparticules d'or. / Improving patients therapeutic care needs the development of sensitive and specific diagnostic tools. Aptamers are synthetic oligonucleotides obtained by SELEX with a very high affinity and excellent specificity for their targets. Grafting of these molecular recognition patterns onto nanomaterials such as gold nanoparticles (GNPs), which unique optical and electronic properties, can amplify the signal induce by the interaction between the ligand and its target. Two biosensing systems have been developed by GNP functionalization with aptamers, one is directed against thrombin and the second against an epigenetic mark carried by a histone protein. Gold precursors was reduced in the presence of PEO4 and led to a homodisperse population of spherical GNP with an average diameter of 14 nm and an isotropy of size and shape. GNP were functionalized with tetraethylene glycol units interconnected by ether or triazoles bridges as a linker. Lipoic acid was used as an anchor moiety onto gold surface via a covalent Au-S bond and was coupled to the spacer through a Steglich reaction. The linkers were functionalized with an azide group to perform the coupling with aptamers by click chemistry. The thrombin sensing strategy used the fluorescence quenching properties of GNPs while the histone detection involved the gold nanoparticle plasmon resonance surface effect.

Made available in DSpace on 2016-04-28T19:59:11Z (GMT). No. of bitstreams: 1 Rosemary Aparecida de Almeida Moraes.pdf: 425632 bytes, checksum: d8599debbfb14f23b04282dc1dc107f5 (MD5) Previous issue date: 2008-05-21 / The purpose of this study is to investigate how the authoral division is processed in the romance Memórias Póstumas de Brás Cubas and what it represents to the realistic-cienficist literary patterns of the XIX century. Having as a central focus the questioning about the authorial signature, the critical analysis concentrated itself in the strategies of this authorial dissemination through three basic principles: the duplicity of the opening prologues between Brás Cubas supposed author and Machado de Assis the authorial signature inscribed in the book; the fiction of the dead author and the conception of autobiography as a portrait of the real author, questioning the frontiers between the fictional and empirical author; the truncated citations and the continuous dislocations of the supposed author Brás Cubas to the function of reader of his own book, which turns the reader into an author, promoting a corrosion of the authorial unity. As a conclusion, the study demonstrated that the questioning of the authorial signature in Memórias Póstumas was an instabilizing element of the romance standard in the realism-naturalism of the XIX century, also affecting the criticism of the sources by means of the authorial reading fiction responsible for the gnawing citations of an author-reader that speaks to us from the other side of the mistery / A proposta deste estudo é investigar como a divisão autoral é processada no romance Memórias Póstumas de Brás Cubas e qual sua representação para o cientificismo-realista literário do século XIX. Apresenta como foco central o questionamento sobre a paternidade autora. A análise crítica se concentrou nas estratégias dessa disseminação autoral por meio de três princípios básicos: a duplicidade dos prólogos de abertura entre Brás Cubas autor suposto - e Machado de Assis - a assinatura autoral inscrita no livro; a ficção do autor morto e a concepção de autobiografia como retrato do autor real, colocando em questão as fronteiras entre autor ficcional e empírico; as citações truncadas e os contínuos deslocamentos do autor suposto Brás Cubas - para a função de leitor do próprio livro que escreve, o que faz do leitor autor, promovendo a corrosão da unicidade autoral. Como conclusão, o estudo demonstrou que o questionamento da paternidade autoral em Memórias Póstumas foi um elemento desestabilizador do modelo de romance do realismo-naturalismo do século XIX, bem como abalou os pressupostos da crítica de fontes por meio da ficção de leitura autoral responsável por citações roídas de um autor-leitor que nos fala do outro lado do mistério

A presente dissertação consiste na comparação dos romances Memórias póstumas de Brás Cubas e Confissões do impostor Felix Krull. Procuramos analisar os aspectos que os aproximam, apesar do afastamento temporal e cultural entre eles existente, partindo dos elementos que imitam o texto autobiográfico e suas respectivas implicações. Tanto as vidas representadas nos dois romances afastam-se do que se entende por uma biografia normal, como também a particular tarefa autobiográfica realizada por Felix e Brás foge do padrão de confissões e de memórias. A confiabilidade dos narradores pode ser ainda questionada, haja vista suas singulares condições enquanto narradores: um, impostor, e outro, defunto. Esta dissertação é composta por seis partes, nas quais se discutirão os seguintes aspectos: introdução à análise comparativa, os romances e os narradores, a intertextualidade inserida na forma e na temática, o decurso de vida de Brás e Felix, suas condições como autobiógrafos e o papel do leitor. No que diz respeito a este último aspecto, podemos afirmar que como o romance em geral, Memórias póstumas e Confissões, que simulam o relato autobiográfico, se preocupam com a função do público. Nos dois livros em questão, essa preocupação reflete-se nos papéis do leitor inseridos nas narrativas. Nossa dissertação ocupar-se-á desse aspecto como último e com particular atenção, visando analisar as reflexões metanarrativas realizadas pelos narradores diante de seus respectivos leitores no horizonte da autobiografia ficcional. / The following degree thesis consists in comparing the novels Memórias póstumas de Brás Cubas and Confissões do impostor Felix Krull. We attempted to analyse the aspects, which approach them, despite of the temporal and cultural distance between them, initiating from the elements, which imitate the autobiographical text and its respective implications. The lives represented in both novels diverge from what usually can be understood as a normal biography. In a similar way the autobiographical function assumed by Felix and Brás diverges from the ideal standard of the confession and the autobiographical memory. The reliability of the narrators can be also questioned if we considerate their singular conditions as narrators: one, as impostor, the other, as a dead person. This thesis is composed of six parts, in that we will discuss the following aspects: an introduction to the comparative analysis, the novels and their narrators, the intertextuality introduced in the form and in the subject, Brás and Felix life events, their condition as autobiographers and the readers role. Like the novels in general, Memórias póstumas e Confissões, that simulate the autobiographical narrative, attend to the role of the attendance. In both analised books this concern is reflected on the the readers inserted in the narratives. This dissertation handles this aspect lastly, trying to analyse the Felix and Brás metanarrative considerations according to their respective readers in the horizon of the fictional autobiography.

Cette thèse s'inscrit dans les domaines de l'apprentissage statistique et de la statistique séquentielle. Le cadre principal est celui des problèmes de bandit stochastique à plusieurs bras. Dans une première partie, on commence par revisiter les bornes inférieures sur le regret. On obtient ainsi des bornes non-asymptotiques dépendantes de la distribution que l'on prouve de manière très simple en se limitant à quelques propriétés bien connues de la divergence de Kullback-Leibler. Puis, on propose des algorithmes pour la minimisation du regret dans les problèmes de bandit stochastique paramétrique dont les bras appartiennent à une certaine famille exponentielle ou non-paramétrique en supposant seulement que les bras sont à support dans l'intervalle unité, pour lesquels on prouve l'optimalité asymptotique (au sens de la borne inférieure de Lai et Robbins) et l'optimalité minimax. On analyse aussi la complexité pour l'échantillonnage séquentielle visant à identifier la distribution ayant la moyenne la plus proche d'un seuil fixé, avec ou sans l'hypothèse que les moyennes des bras forment une suite croissante. Ce travail est motivé par l'étude des essais cliniques de phase I, où l'hypothèse de croissance est naturelle. Finalement, on étend l'inégalité de Fano qui contrôle la probabilité d'évènements disjoints avec une moyenne de divergences de Kullback-leibler à des variables aléatoires arbitraires bornées sur l'intervalle unité. Plusieurs nouvelles applications en découlent, les plus importantes étant une borne inférieure sur la vitesse de concentration de l'a posteriori Bayésien et une borne inférieure sur le regret pour un problème de bandit non-stochastique. / The topics addressed in this thesis lie in statistical machine learning and sequential statistic. Our main framework is the stochastic multi-armed bandit problems. In this work we revisit lower bounds on the regret. We obtain non-asymptotic, distribution-dependent bounds and provide simple proofs based only on well-known properties of Kullback-Leibler divergence. These bounds show in particular that in the initial phase the regret grows almost linearly, and that the well-known logarithmic growth of the regret only holds in a final phase. Then, we propose algorithms for regret minimization in stochastic bandit models with exponential families of distributions or with distribution only assumed to be supported by the unit interval, that are simultaneously asymptotically optimal (in the sense of Lai and Robbins lower bound) and minimax optimal. We also analyze the sample complexity of sequentially identifying the distribution whose expectation is the closest to some given threshold, with and without the assumption that the mean values of the distributions are increasing. This work is motivated by phase I clinical trials, a practically important setting where the arm means are increasing by nature. Finally we extend Fano's inequality, which controls the average probability of (disjoint) events in terms of the average of some Kullback-Leibler divergences, to work with arbitrary unit-valued random variables. Several novel applications are provided, in which the consideration of random variables is particularly handy. The most important applications deal with the problem of Bayesian posterior concentration (minimax or distribution-dependent) rates and with a lower bound on the regret in non-stochastic sequential learning.

Dans cette thèse, nous étudions des problèmes de prise de décisions séquentielles dans lesquels, pour chacune de ses décisions, l'apprenant reçoit une information qu'il utilise pour guider ses décisions futures. Pour aller au-delà du retour d’information conventionnel tel qu'il a été bien étudié pour des problèmes de prise de décision séquentielle tels que les bandits multi-bras, nous considérons des formes de retour d’information partielle motivées par des applications pratiques.En premier, nous considérons le problème des bandits duellistes, dans lequel l'apprenant sélectionne deux actions à chaque pas de temps et reçoit en retour une information relative (i.e. de préférence) entre les valeurs instantanées de ces deux actions.En particulier, nous proposons un algorithme optimal qui permet à l'apprenant d'obtenir un regret cumulatif quasi-optimal (le regret est la différence entre la récompense cumulative optimale et la récompense cumulative constatée de l’apprenant). Dans un second temps, nous considérons le problème des bandits corrompus, dans lequel un processus de corruption stochastique perturbe le retour d’information. Pour ce problème aussi, nous concevons des algorithmes pour obtenir un regret cumulatif asymptotiquement optimal. En outre, nous examinons la relation entre ces deux problèmes dans le cadre du monitoring partiel qui est un paradigme générique pour la prise de décision séquentielle avec retour d'information partielle. / The multi-armed bandit (MAB) problem is a mathematical formulation of the exploration-exploitation trade-off inherent to reinforcement learning, in which the learner chooses an action (symbolized by an arm) from a set of available actions in a sequence of trials in order to maximize their reward. In the classical MAB problem, the learner receives absolute bandit feedback i.e. it receives as feedback the reward of the arm it selects. In many practical situations however, different kind of feedback is more readily available. In this thesis, we study two of such kinds of feedbacks, namely, relative feedback and corrupt feedback.The main practical motivation behind relative feedback arises from the task of online ranker evaluation. This task involves choosing the optimal ranker from a finite set of rankers using only pairwise comparisons, while minimizing the comparisons between sub-optimal rankers. This is formalized by the MAB problem with relative feedback, in which the learner selects two arms instead of one and receives the preference feedback. We consider the adversarial formulation of this problem which circumvents the stationarity assumption over the mean rewards for the arms. We provide a lower bound on the performance measure for any algorithm for this problem. We also provide an algorithm called "Relative Exponential-weight algorithm for Exploration and Exploitation" with performance guarantees. We present a thorough empirical study on several information retrieval datasets that confirm the validity of these theoretical results.The motivating theme behind corrupt feedback is that the feedback the learner receives is a corrupted form of the corresponding reward of the selected arm. Practically such a feedback is available in the tasks of online advertising, recommender systems etc. We consider two goals for the MAB problem with corrupt feedback: best arm identification and exploration-exploitation. For both the goals, we provide lower bounds on the performance measures for any algorithm. We also provide various algorithms for these settings. The main contribution of this module is the algorithms "KLUCB-CF" and "Thompson Sampling-CF" which asymptotically attain the best possible performance. We present experimental results to demonstrate the performance of these algorithms. We also show how this problem setting can be used for the practical application of enforcing differential privacy.

Submitted by Irvana Coutinho (irvana@ufpa.br) on 2017-04-17T16:55:07Z No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Dissertacao_SubsoloNilismoLiteratura.pdf: 1142767 bytes, checksum: 4c977fdc11af76214acc93a21366e2ff (MD5) / Approved for entry into archive by Irvana Coutinho (irvana@ufpa.br) on 2017-04-17T16:55:21Z (GMT) No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Dissertacao_SubsoloNilismoLiteratura.pdf: 1142767 bytes, checksum: 4c977fdc11af76214acc93a21366e2ff (MD5) / Made available in DSpace on 2017-04-17T16:55:21Z (GMT). No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Dissertacao_SubsoloNilismoLiteratura.pdf: 1142767 bytes, checksum: 4c977fdc11af76214acc93a21366e2ff (MD5) Previous issue date: 2016-03-28 / Machado de Assis é, com justeza, um grande escritor brasileiro não somente pela sua criatividade e escrita, mas principalmente pelas suas ideias, sua postura diante dos valores, que não raro são questionados em suas obras, o que poderia representar uma postura que, longe de ser simplesmente pessimista, está voltada mais para um padrão niilista de pensamento e visão de mundo. Questionar nem sempre significa destruir. Mas se a destruição se faz inevitável, há de se repor outros valores. A morte nem sempre se configura como um fim, mas pode claramente representar o inicio de uma nova existência. Neste trabalho, será observada a questão do niilismo em Memórias Póstumas de Brás Cubas, acionando filósofos como Nietzsche, para pô-los em diálogo com outros escritores, como Dostoiévski, demonstrando que o niilismo não pode ser considerado um fenômeno regional, mas mundial, demonstrando que em Brás Cubas, tal niilismo converge para outros padrões e para uma possibilidade de resistência através do humor, do riso, da alegria. / Machado de Assis is, rightly, a great Brazilian writer not only for their creativity and writing, but mostly for his ideas, his attitude toward the values that often are questioned in his works, which could represent a position that far to be just pessimistic, it is geared more to a nihilist pattern of thought and worldview. Question does not always mean destroy. But if the destruction is inevitable if there is to reset other values. Death is not always configured as an end, but can clearly represent the beginning of a new existence. In this work, it will be subject to the issue of nihilism in The Posthumous Memoirs of Bras Cubas, triggering philosophers like Nietzsche, to put them in dialogue with other writers like Dostoevsky, demonstrating that nihilism can not be considered a regional phenomenon, but worldwide, demonstrating that in Brás Cubas, such nihilism converges to other standards and the possibility of resistance through humor, laughter, joy.

Cette thèse comprend deux parties relativement distinctes. Dans la première, nous présentons l'analyse de 16 millions de courbes de lumière collectées par EROS II vers différentes longitudes dans le plan Galactique durant 7 saisons d'observation. Après avoir crée des images de référence et des courbes de lumière, nous avons recherché des effets de microlentille gravitationnelle parmi les courbes de lumière produites en appliquant une analyse discriminante, tout en calculant notre efficacité de détection sur un lot de courbes de lumière simulées. Nous avons mis en évidence 24 candidats, ce qui nous a permis de mesurer la profondeur optique vers nos cibles. La profondeur optique moyenne vers les bras spiraux est $<\tau_(GSA)> = 0.38\pm0.08$. Notre résultat est compatible avec les prédictions de modèles Galactiques tirés de la littérature, dans lesquels le bulbe central est modélisé par une barre, et favorise ceux sans matière cachée dans le disque. Outre la statistique, la principale limi tation dans notre analyse provient du fait que les distances où sont situées les étoiles cibles sont mal connues.<br> Dans la seconde partie, nous présentons l'analyse des données spectroscopiques, collectées durant une campagne internationale de recherche de supernovae Ia menée au printemps 1999, à laquelle EROS II a participé. Une vingtaine de supernovae Ia ont été découvertes puis suivies pendant cette campagne. Nous avons réduit les données spectroscopiques les concernant, qui consistent en une centaine de spectres au total. Nous avons developpé pour celà un programme original permettant de séparer le flux de la supernovae de celui de sa galaxie hôte. Nous avons ensuite procédé à une étude de stan dardisations sur un lot regroupant des objets de la campagne 1999 et des objets découverts antérieuement. L'étude sur ces 17 supernovae Ia a permis de confirmer l'interêt des spectres pour la standardis ation. Elle montre en particulier que la standardisation à partir de spectres est aussi efficace que celle basée sur le taux de décroissance des courbes de lumière.

En pratique clinique, les résultats de l’Anesthésie Loco-Régionale diffèrent parfois notablement de ceux prévus par les traités d’anatomie. Le but de notre travail a été de montrer que les nerfs musculocutané (MC) et cutanémédial de l’avant-bras (CMAB) participent fréquemment à l’innervation cutanée de la paume de la main. L’analyse multivariée de 551 patients opérés du canal carpien a révélé que l’absence de bloc du nerf MC était associée aux échecs de l’ALR. Pourtant, ce résultat était en partie masqué par une fréquente diffusion d’anesthésique local du nerf médian vers le nerf MC au bras, comme le prouve l’étude échographique de l’anatomie des nerfs au creux axillaire chez 387 autres patients. Une cartographie du territoire cutané des nerfs MC (N=28) et CMAB (N=2l) à la face antérieure de la main et du poignet a révélé des territoires bien plus étendus que ceux décrits dans la littérature. Les microdissections de ces 2 nerfs sur 23 membres supérieurs ont permis d’associer les différents types de territoires décrits lors des cartographies à diverses configurations anatomiques. Il convient donc d’anesthésier ces deux nerfs pour toute chirurgie de la paume de la main. / In daily practice, the extent of peripheral nerve blockade often differs from the one predicted by referencetextbooks. In this work, we strived to demonstrate that musculocutaneous (MC) and medial antebrachialcutaneous (MABC) nerves participate frequently in the palm innervation. A multivariate analysis of 551 patients operated from carpal tunnel release showed that the absence of MC nerve block was associated with anesthesia failure. Yet, these results were minimized by frequent local anesthetic diffusion from the median nerve to the MC one at the arm, as demonstrated by our anatomical study of the nerves in 387 ultrasound guided axillary blocks. Cutaneous territory mapping of MC (N28) and MABC (N=2 1) nerves revealed their territories were far more extended than the ones described in reference textbooks. Microdissections of these nerves on 23 upper limbs allow to associate the different types of territories with several anatomical patterns. Therefore, MC and MABC nerve should be blocked when considering any palm surgery.