Cette thèse est consacrée à l'étude des valeurs propres de l'opérateur de Laplace et de l'opérateur de Steklov sur des variétés riemanniennes. On cherche à donner des bornes optimales parmi l'ensemble des métriques, dans une classe conforme donnée ou non, et à caractériser, si elles existent, les métriques qui atteignent ces bornes. Ces métriques extrémales ont des propriétés qui s'inscrivent dans la théorie des surfaces minimales. On s'intéresse d'abord à la borne supérieure des valeurs propres de Laplace parmi des métriques conformes entre elles, appelées valeurs propres conformes. Dans le chapitre 1, on estime la deuxième valeur propre conforme de la sphère standard. Dans les chapitres 2 et 3, on montre que la première valeur propre conforme d'une variété riemannienne est plus grande que celle de la sphère standard de même dimension avec égalité seulement pour la sphère standard. Ensuite, on cherche à démontrer l'existence et la régularité de métriques qui maximisent les valeurs propres sur des surfaces, dans une classe conforme donnée ou non. Dans les chapitres 3 et 4, on démontre un résultat d'existence pour les valeurs propres de Laplace. Dans le chapitre 6, le travail est fait pour les valeurs propres de Steklov. Enfin, dans le chapitre 5, fruit d'un travail réalisé en collaboration avec Paul Laurain, on démontre un résultat de régularité et de quantification des applications harmoniques à bord libre sur une surface Riemannienne. C'est un élément clé pour le chapitre 6 / This thesis is devoted to the study of the Laplace eigenvalues and the Steklov eigenvalues on Riemannian manifolds. We look for optimal bounds among the set of metrics, lying in a conformal class or not. We also characterize, if they exist the metrics which reach these bounds. These extremal metrics have properties from the theory of minimal surfaces. First, we are interested in the upper bound of Laplace eigenvalues in a class of conformal metrics, called the conformal eigenvalues. In Chapter 1, we estimate the second conformal eigenvalue of the standard sphere. In Chapters 2 and 3, we prove that the first conformal eigenvalue of a Riemannian manifold is greater than the one of the standard sphere of same dimension, with equality only for the standard sphere. Then, we look for existence and regularity results for metrics which maximize eigenvalues on surfaces, in a given conformal class or not. In Chapters 3 and 4, we prove an existence result for Laplace eigenvalues. In Chapter 6, the work is done for Steklov eigenvalues. Finally, in Chapter 5, obtained in collaboration with Paul Laurain, we prove a regularity and quantification result for harmonic maps with free boundary on a Riemannian surface. It is a key component for Chapter 6

As teorias científicas surgiram da necessidade do homem entender o funcionamento das coisas. Novos métodos e técnicas são então criados com o objetivo não só de melhor compreender, mas também de desenvolver essas próprias teorias. Nesta dissertação, vamos estudar várias dessas técnicas (aqui chamadas de algoritmos) com o objetivo de obter estados fundamentais em sistemas de spin e de revelar suas possíveis propriedades de auto-organização crítica. No segundo capítulo desta dissertação, apresentamos os algoritmos de otimização: simulated annealing, algoritmo genético, otimização extrema (EO) e evolutivo de Bak-Sneppen (BS). No terceiro capítulo apresentamos o conceito de criticalidade auto-organizada (SOC), usando como exemplo o modelo da pilha de areia. Para uma melhor compreensão da importância da criticalidade auto-organizada, apresentamos vários outros exemplos de onde o fenômeno é observado. No quarto capítulo apresentamos o modelo de relógio quiral de p-estados que será nosso sistema de testes. No caso unidimensional, determinamos a matriz de transferência e utilizamos o teorema de Perron-Frobenius para provar a inexistência de transição de fase a temperaturas finitas a temperaturas finitas. Esboçamos os diagramas de fases dos estados fundamentais que obtivemos de maneira analítica e numérica para os casos de p = 2, 3, 4, 5 e 6, no caso numérico fazendo uso do algoritmo de Bak-Sneppen com sorteio (BSS). Apresentamos ainda um breve estudo do número de mínimos locais para o modelo de relógio quiral de p-estados, para os casos de p = 3 e 4. Por último, no quinto capítulo, propomos uma dinâmica Bak-Sneppen com ruído (BSR) como uma nova técnica de otimização para tratar sistemas discretos. O ruído é introduzido diretamente no espaço de configuração de spins. Conseqüentemente, o fitness (adaptabilidade) passa a assumir valores contínuos, num pequeno intervalo em torno do seu valor original (discreto). Os resultados dessa dinâmica indicam a presença de criticalidade auto-organizada, evidenciada pelo decaimento em leis de potências das correlações espacial e temporal. Também estudamos o método EO e obtivemos uma confirmação numérica de que sua dinâmica exibe um comportamento não crítico com alcance espacial infinito e decaimento exponencial das avalanches. Finalmente, para o modelo de relógio quiral, comparamos a eficiência das três dinâmicas (EO, BSS e BSR) no que tange às suas habilidades de encontrar o estado fundamental do sistema. / In order to understand how things work, man has formulated scientific theories. New methods and techniques have been created not only to increase our understanding on the subject but also to develop and even expand those theories. In this thesis, we study several techniques (here called algorithms) designed with the objective to get the ground states of some spin systems and eventually to reveal possible properties of critical self-organization. In the second chapter, we introduce four fundamental optimization algorithms: simulated annealing, genetics algorithms, extremal optimization (EO) and Bak-Sneppen (BS). In the third chapter we present the concept of self-organized criticality (SOC), using as an example the sandpile model. To understand the importance of the self-organized criticality, we show many other situations where the phenomenon can be observed. In the fourth chapter, we introduce the p-states chiral clock model. This will be our test or toy system. For the one-dimensional case, we first determined the corresponding transfer-matrix and then proved the nonexistence of phase transitions by using the Perron-Frobenius theorem. We calculate the ground state phase diagrams both analytically and numerically in the cases of p = 2, 3, 4, 5 and 6. We also present a brief study of the number of local minima for the cases p = 3 and 4 of the chiral clock model. Finally, in the fifth chapter, we propose a Bak-Sneppen dynamics with noise (BSN) as a new technique of optimization to treat discrete systems. The noise is directly introduced into the spin configuration space. Consequently, the fitness now take values in a continuum but small interval around its original value (discrete). The results of this dynamics indicate the presence of self-organized criticality, which becomes evident with the power law scaling of the spacial and temporal correlations. We also study the EO algorithm and found a numerical con_rmation that it does not show a critical behavior since it has an in_nite space range and an exponential decay of the avalanches. At the end, we compare the e_ciency of the three dynamics (EO, BSD and BSN) for the chiral clock model, concerning their abilities to _nd the system\'s ground state.

In this thesis are extreme value theory used to estimate the probability that large insuranceclaims are exceeding a certain threshold. The expected claim size, given that the claimhas exceeded a certain limit, are also estimated. Two different models are used for thispurpose. The first model is based on maximum domain of attraction conditions. A Paretodistribution is used in the other model. Different graphical tools are used to check thevalidity for both models. Länsförsäkring Kronoberg has provided us with insurance datato perform the study.Conclusions, which have been drawn, are that both models seem to be valid and theresults from both models are essential equal. / I detta arbete används extremvärdesteori för att uppskatta sannolikheten att stora försäkringsskadoröverträffar en vis nivå. Även den förväntade storleken på skadan, givetatt skadan överstiger ett visst belopp, uppskattas. Två olika modeller används. Den förstamodellen bygger på antagandet att underliggande slumpvariabler tillhör maximat aven extremvärdesfördelning. I den andra modellen används en Pareto fördelning. Olikagrafiska verktyg används för att besluta om modellernas giltighet. För att kunna genomförastudien har Länsförsäkring Kronoberg ställt upp med försäkringsdata.Slutsatser som dras är att båda modellerna verkar vara giltiga och att resultaten ärlikvärdiga.

In 1988, Owen introduced empirical likelihood as a nonparametric method for constructing confidence intervals and regions. Since then, empirical likelihood has been studied extensively in the literature due to its generality and effectiveness. It is well known that empirical likelihood has several attractive advantages comparing to its competitors such as bootstrap: determining the shape of confidence regions automatically using only the data; straightforwardly incorporating side information expressed through constraints; being Bartlett correctable. The main part of this thesis extends the empirical likelihood method to several interesting and important statistical inference situations. This thesis has four components. The first component (Chapter II) proposes a smoothed jackknife empirical likelihood method to construct confidence intervals for the receiver operating characteristic (ROC) curve in order to overcome the computational difficulty when we have nonlinear constrains in the maximization problem. The second component (Chapter III and IV) proposes smoothed empirical likelihood methods to obtain interval estimation for the conditional Value-at-Risk with the volatility model being an ARCH/GARCH model and a nonparametric regression respectively, which have applications in financial risk management. The third component(Chapter V) derives the empirical likelihood for the intermediate quantiles, which plays an important role in the statistics of extremes. Finally, the fourth component (Chapter VI and VII) presents two additional results: in Chapter VI, we present an interesting result by showing that, when the third moment is infinity, we may prefer the Student's t-statistic to the sample mean standardized by the true standard deviation; in Chapter VII, we present a method for testing a subset of parameters for a given parametric model of stationary processes.

This paper links two applications of Extreme Value Theory (EVT) to analyze Taiwanese financial markets: 1. computation of Value at Risk (VaR) and Expected Shortfall (ES) 2. estimates of cross-market dependence under extreme events. Daily data from the Taiwan Stock Exchange Capitalization Weight Stock Index (TAIEX) and the foreign exchange rate, USD/NTD, are employed to analyze the behavior of each return and the dependence structure between the foreign exchange market and the equity market. In the univariate case, when computing risk measures, EVT provides us a more accurate way to estimate VaR. In bivariate case, when measuring extremal dependence, the results of whole period data show the extremal dependence between two markets is asymptotically independent, and the analyses of subperiods illustrate that the relation is slightly dependent in specific periods. Therefore, there is no significant evidence that extreme events appeared in one market (the equity market or the foreign exchange market) will affect another in Taiwan.

In dieser Arbeit wird eine Beschreibung von Monte-Carlo-Verfahren zur Lösung komplexer Optimierungsaufgaben mit Hilfe von Markov-Ketten durchgeführt. Nach einer kurzen Einführung werden Lösungsmenge solcher Aufgaben und der physikalische Zustandsraum komplexer Systeme identifiziert. Zunächst wird die Dynamik von Zufallswanderern im Zustandsraum mit Hilfe von Master-Gleichungen modelliert. Durch Einführung von Performanzkriterien können verschiedene Optimierungsstrategien quantitativ miteinander verglichen werden. Insbesondere wird das Verfahren Extremal Optimization vorgestellt, dass ebenfalls als Markov-Prozess verstanden werden kann. Es wird bewiesen, dass eine im Sinne der genannten Kriterien beste Implementierung existiert. Da diese von einem sogenannten Fitness Schedule abhängt, wird dieser für kleine Beispielsysteme explizit berechnet. Daran anschließend wird die Zustandsdichte komplexer Systeme betrachtet. Nach einem kurzen Überblick über vorhandene Methoden folgt eine detaillierte Untersuchung des Verfahrens von Wang und Landau. Numerische und analytische Hinweise werden gegeben, nach denen dieser Algorithmus innerhalb seiner Klasse wahrscheinlich der Optimale ist. Eine neue Methode zur Approximation der Zustandsdichte wird vorgestellt, die insbesondere für die Untersuchung komplexer Systeme geeignet ist. Abschließend wird ein Ausblick auf zukünftige Arbeiten gegeben.

No description available.

Understanding the behavior of complex, large-scale, interconnected systems in a rigorous and structured manner is one of the most pressing scientific and technological challenges of current times. These systems include, among many others, transportation and communications systems, smart grids and power grids, financial markets etc. Failures of these systems have potentially enormous social, environmental and financial costs. In this work, we investigate the failure mechanisms of load-sharing complex systems. The systems are composed of multiple nodes or components whose failures are determined based on the interaction of their respective strengths and loads (or capacity and demand respectively) as well as the ability of a component to share its load with its neighbors when needed. Each component possesses a specific strength (capacity) and can be in one of three states: failed, damaged or functioning normally. The states are determined based on the load (demand) on the component. We focus on two distinct mechanisms to model the interaction between components strengths and loads. The first, a Loss of Strength (LOS) model and the second, a Customer Service (CS) model. We implement both models on lattice and scale-free graph network topologies. The failure mechanisms of these two models demonstrate temporal scaling phenomena, phase transitions and multiple distinct failure modes excited by extremal dynamics. We find that the resiliency of these models is sensitive to the underlying network topology. For critical ranges of parameters the models demonstrate power law and exponential failure patterns. We find that the failure mechanisms of these models have parallels to failure mechanisms of critical infrastructure systems such as congestion in transportation networks, cascading failure in electrical power grids, creep-rupture in composite structures, and draw-downs in financial markets. Based on the different variants of failure, strategies for mitigating and postponing failure in these critical infrastructure systems can be formulated.

Dans ce travail nous nous intéressons à la géométrie de l’espace de Teichmüller via la longueur extrémale et à sa relation avec d’autres géométries. En effet, via le théorème d’uniformisation de Poincaré, l’espace de Teichmüller d’une surface orientable de type finie est un espace qui “classifie” aussi bien les structures hyperboliques de cette surface que les structures conformes. Suivant la classification utilisée, on obtient deux compactifications différentes de cet espace, qui sont respectivement la compactification de Thurston et la compactification de Gardiner-Masur. La première étant induite par la longueur hyperbolique et la deuxième par la longueur extrémale. Dans une première partie, on considère les compactifications dites “réduites” de Thurston et Gardiner-Masur. On montre qu’il existe une bijection naturelle entre les deux et que le groupe des auto-homéomorphismes du bord réduit de Thurston est canoniquement isomorphe au groupe modulaire étendu de la surface sous-jacente. Dans une deuxième partie, on étudie la convergence de certaines déformations de structures conformes aussi bien sur le bord de Thurston que sur celui de Gardiner-Masur. Ces déformations, appelées déformations horocycliques, sont un analogue des tremblements de terre de structures hyperboliques. Enfin, dans une troisième et dernière partie, on introduit une compactification à la Gardiner-Masur de l’espace de Teichmüller d’une surface à bord. On généralise des résultats obtenus dans le cas sans bord, et on établit quelques différences. / In this thesis we are interested in the extremal length geometry of Teichmüller space and the links with other geometries. In particular, we work on two different compactifications of Teichmüller space, namely, the Thurston compactification and the Gardiner-Masur compactification. In the first part, we consider the so-called reduced compactifications of Thurston and Gardiner-Masur. We show that there exists a canonical bijection between them and that the group of self-homeomorphisms of the reduced Thurston boundary is canonicaly isomorphic (except for a few cases) to the extended mapping class group of the corresponding surface. In the second part, we study the asymptotic behaviour of some conformal structure deformations to the Thuston boundary and to the Gardiner-Masur boundary. These deformations are called horocyclic deformations and they are analogous to earthquakes of hyperbolic structures. Finally, in the last part, using extremal length we extend the notion of Gardiner-Masur compactification to surfaces with non-empty boundary, and we investigate differences with the case without boundary.

Cette thèse traite principalement des théorèmes limites pour les processus empiriques de fonctionnelles de clusters d'extrêmes de séquences et champs aléatoires faiblement dépendants. Des théorèmes limites pour les processus empiriques de fonctionnelles de clusters d'extrême de séries temporelles stationnaires sont donnés par Drees & Rootzén [2010] sous des conditions de régularité absolue (ou "ß-mélange"). Cependant, ces conditions de dépendance de type mélange sont très restrictives : elles sont particulièrement adaptées aux modèles dans la finance et dans l'histoire, et elles sont de plus compliquées à vérifier. Généralement, pour d'autres modèles fréquemment rencontré dans les domaines applicatifs, les conditions de mélange ne sont pas satisfaites. En revanche, les conditions de dépendance faible, selon Doukhan and Louhichi [1999] et Dedecker & Prieur [2004a], sont des conditions qui généralisent les notions de mélange et d'association. Elles sont plus simple à vérifier et peuvent être satisfaites pour de nombreux modèles. Plus précisément, sous des conditions faibles, tous les processus causals ou non causals sont faiblement dépendants: les processus Gaussien, associés, linéaires, ARCH(∞), bilinéaires et notamment Volterra entrent dans cette liste. À partir de ces conditions favorables, nous étendons certains des théorèmes limites de Drees & Rootzén [2010] à processus faiblement dépendants. En outre, comme application des théorèmes précédents, nous montrons la convergence en loi de l'estimateur de l'extremogramme de Davis & Mikosch [2009] et l'estimateur fonctionnel de l'indice extrémal de Drees [2011] sous dépendance faible. Nous démontrons un théorème de la valeur extrême pour les champs aléatoires stationnaires faiblement dépendants et nous proposons, sous les mêmes conditions, un critère du domaine d'attraction d'une loi d'extrêmes. Le document se conclue sur des théorèmes limites pour les processus empiriques de fonctionnelles de clusters d’extrêmes de champs aléatoires stationnaires faiblement dépendants, et met en évidence la convergence en loi de l'estimateur d'un extremogramme de processus spatio-temporels stationnaires faiblement dépendants en tant qu'application. / This thesis deals mainly with limit theorems for empirical processes of extreme cluster functionals of weakly dependent random fields and sequences. Limit theorems for empirical processes of extreme cluster functionals of stationnary time series are given by Drees & Rootzén [2010] under absolute regularity (or "ß-mixing") conditions. However, these dependence conditions of mixing type are very restrictive: on the one hand, they are best suited for models in finance and history, and on the other hand, they are difficult to verify. Generally, for other models common in applications, the mixing conditions are not satisfied. In contrast, weak dependence conditions, as defined by Doukhan & Louhichi [1999] and Dedecker & Prieur [2004a], are dependence conditions which generalises the notions of mixing and association. These are easier to verify and applicable to a wide list of models. More precisely, under weak conditions, all the causal or non-causal processes are weakly dependent: Gaussian, associated, linear, ARCH(∞), bilinear and Volterra processes are some included in this list. Under these conveniences, we expand some of the limit theorems of Drees & Rootzén [2010] to weakly dependent processes. These latter results are used in order to show the convergence in distribution of the extremogram estimator of Davis & Mikosch [2009] and the functional estimator of the extremal index introduced by Drees [2011] under weak dependence. We prove an extreme value theorem for weakly dependent stationary random fields and we propose, under the same conditions, a domain of attraction criteria of a law of extremes. The document ends with limit theorems for the empirical process of extreme cluster functionals of stationary weakly dependent random fields, deriving also the convergence in distribution of the estimator of an extremogram for stationary weakly dependent space-time processes.