Die Dissertation mit dem Titel “Battle and Beating, Water and Waste: Micro-Level Impact Evaluation in Developing and Emerging Economies” beinhaltet die Ergebnisse ökonometrischer Wirkungsevaluierungen, die innerhalb zwei verschiedener Themengebieten des täglichen Lebens in Entwicklungs- und Schwellenländern durchgeführt wurden: Gewalt und Wasser. Im ersten dieser Themengebiete konzentriert sich die Analyse auf die Verhaltens-Reaktionen der Bevölkerung, die in räumlicher Nähe zu gewaltsamem Konflikt lebt. Detailliert die Wirkung solcher Exponierung auf das Auftreten häuslicher Gewalt untersuchend, trägt das erste Kapitel – gemeinsam verfasst mit Dominik Noe – dazu bei, die gesellschaftlichen Kosten kriegsähnlicher Auseinandersetzungen abzuschätzen. Insbesondere die Nachwirkungen auf zwischenmenschliche Beziehungen und soziales Verhalten gegenüber Familie und Freunden werden betrachtet. Wir formulieren die Theorie, dass das Leben in Haushalten in der Nähe zu Orten von extrem gewaltsamen Zwischenfällen die Wahrscheinlichkeit von Frauen in diesen Haushalten erhöht, Opfer häuslicher Gewalt zu werden. Diese Theorie überprüfen wir dann anhand von Daten aus Kolumbien; einem Land, in dem sowohl reichhaltige Daten zu Gefechten als auch zu häuslicher Gewalt vorliegen. Die Kernerkenntnis lautet, dass eine höhere Intensität der gewaltsamen Konflikte mutmaßlich die Wahrscheinlichkeit der Frauen, Opfer häuslicher Gewalt zu werden, deutlich erhöht. Das zweite und dritte Kapitel dieser Dissertation sind dem zweiten Themengebiet gewidmet, und hier insbesondere Aspekten des Anschlusses von Haushalten an Leitungswasser und moderne Abwasser-Entsorgung. Das zweite Kapitel – gemeinsam verfasst mit Stephan Klasen, Tobias Lechtenfeld und Kristina Meier – untersucht die Wirkung des Anschlusses von Haushalten an Leitungswasser und Abwasser-entsorgung auf Gesundheit sowie Schul- und Arbeitsplatz-Anwesenheit. Es trägt zu der derzeit noch überschaubaren Anzahl von Wirkungsevaluierungen im Wasser- und Abwasserbereich bei, und ist nach Kenntnis der Autoren die erste rigorose Wirkungsevaluierung im städtischen Umfeld. Quasi-experimentelle Methoden und Wasserqualitäts-Tests unter Berücksichtigung von ungleichmäßiger Infrastruktur-Ausbreitung erlauben es, die Wirkungen separat einerseits für Anschluss lediglich an Leitungswasser, als auch andererseits für zusätzlichen Anschluss an Abwasserentsorgung, zu schätzen. Die Kernerkenntnis lautet, dass Anschluss an Leitungswasser in jemenitischen Städten schädlich sein kann, wenn die Wasserversorgung unregelmäßig ist; und die Wirkung bei regelmäßiger Versorgung im Vergleich zu traditionellen und alternativen Wasserquellen sehr gering ist. Anschluss an Abwasserentsorgung scheint dagegen bei regelmäßiger Wasserversorgung zu gesundheitlichem Nutzen zu führen. Es ist regelmäßig zu beobachten, dass in Aufbereitungsanlagen entkeimtes Leitungswasser innerhalb des Haushaltes rekontaminiert wird. Wasserqualitäts-Tests an Testpunkten entlang der Versorgungskette innerhalb des Haushaltes erlaubt Rückschlüsse auf Ort und Quelle der Kontaminierung. Der Löwenanteil dieser Verschlechterung der Wasserqualität ist auf Verhalten im Zusammenhang mit Wasser-Handhabung und Hygiene zurückzuführen. Das in Alleinautorschaft verfasste dritte Kapitel baut auf dem zweiten auf und untersucht, welche treibenden Faktoren hinter Verhaltensaspekten stehen, die Rekontaminierung verhindern würden. Spezifisch die Determinanten von Wasserbehandlung und Hygiene werden untersucht, und vor dem Hintergrund der Erkenntnisse aus dem zweiten Kapitel interpretiert. Die Kernerkenntnis lautet, dass Hygiene-Training, Zugang zu Informations- und Kommunikations-Technologie sowie Schulbildung mutmaßlich zu den relevanten, beeinflussbaren Determinanten zählen. Anschluss an Leitungswasser und Abwasserentsorgung – welche üblicherweise bestenfalls mittelfristig ausgebaut werden können – haben scheinbar ebenfalls wünschenswerte Effekte, allerdings in geringerem Maße, und in höherem Maße bei Leitungswasser als bei zusätzlicher Abwasserentsorgung.

Quand on veut évaluer ou manipuler une fonction mathématique f, il est fréquent de la remplacer par une approximation polynomiale p. On le fait, par exemple, pour implanter des fonctions élémentaires en machine, pour la quadrature ou la résolution d'équations différentielles ordinaires (ODE). De nombreuses méthodes numériques existent pour l'ensemble de ces questions et nous nous proposons de les aborder dans le cadre du calcul rigoureux, au sein duquel on exige des garanties sur la précision des résultats, tant pour l'erreur de méthode que l'erreur d'arrondi.Une approximation polynomiale rigoureuse (RPA) pour une fonction f définie sur un intervalle [a,b], est un couple (P, Delta) formé par un polynôme P et un intervalle Delta, tel que f(x)-P(x) appartienne à Delta pour tout x dans [a,b].Dans ce travail, nous analysons et introduisons plusieurs procédés de calcul de RPAs dans le cas de fonctions univariées. Nous analysons et raffinons une approche existante à base de développements de Taylor.Puis nous les remplaçons par des approximants plus fins, tels que les polynômes minimax, les séries tronquées de Chebyshev ou les interpolants de Chebyshev.Nous présentons aussi plusieurs applications: une relative à l'implantation de fonctions standard dans une bibliothèque mathématique (libm), une portant sur le calcul de développements tronqués en séries de Chebyshev de solutions d'ODE linéaires à coefficients polynômiaux et, enfin, un processus automatique d'évaluation de fonction à précision garantie sur une puce reconfigurable. / For purposes of evaluation and manipulation, mathematical functions f are commonly replaced by approximation polynomials p. Examples include floating-point implementations of elementary functions, integration, ordinary differential equations (ODE) solving. For that, a wide range of numerical methods exists. We consider the application of such methods in the context of rigorous computing, where we need guarantees on the accuracy of the result, with respect to both the truncation and rounding errors.A rigorous polynomial approximation (RPA) for a function f defined over an interval [a,b] is a couple (P, Delta) where P is a polynomial and Delta is an interval such that f(x)-P(x) belongs to Delta, for all x in [a,b]. In this work we analyse and bring forth several ways of obtaining RPAs for univariate functions. Firstly, we analyse and refine an existing approach based on Taylor expansions. Secondly, we replace them with better approximations such as minimax approximations, Chebyshev truncated series or interpolation polynomials.Several applications are presented: one from standard functions implementation in mathematical libraries (libm), another regarding the computation of Chebyshev series expansions solutions of linear ODEs with polynomial coefficients, and finally an automatic process for function evaluation with guaranteed accuracy in reconfigurable hardware.

Bragg scattering of optical waves in thick gratings at extreme angles, where the scattered wave propagates parallel (extremely asymmetric scattering - EAS) or nearly parallel (grazing angle scattering - GAS) to the grating boundaries, is associated with many unique and practically important resonant phenomena. It has been demonstrated that one of the main physical mechanisms for these resonant phenomena is the diffractional divergence of the scattered wave inside and outside the grating region. This thesis fills the gaps in the theoretical and experimental understanding of Bragg scattering in gratings at extreme angles by investigating EAS and GAS in structures where diffractional divergence of waves is significantly affected by anisotropy and/or non-uniformities of the dielectric permittivity. Unusually high sensitivity of wave scattering in thick periodic gratings to small step-like variations of mean structural parameters at the grating boundaries is predicted and described for the case when the scattered wave (the +1 diffracted order) propagates almost parallel to the front grating boundary (the geometry of GAS). A unusual pattern of strong multiple resonances for bulk electromagnetic waves is predicted and analysed numerically in thick periodic holographic gratings in a guiding slab with mean permittivity that is greater than that of the surrounding media. It is demonstrated that these resonances are related to resonant generation of a new type of eigenmodes in a thick slab with a periodic grating. These eigenmodes are generically related to the grating -- they do exist not if the grating amplitude is zero. A new type of resonant coupling of bulk radiation into the conventional guided modes of a slab with a thick holographic grating is predicted and explained theoretically. It occurs in the presence of strong frequency detunings of the Bragg condition by means of interaction of the strongly non-eigen +1 diffracted order with the slab-grating boundaries. Therefore, it is only in the presence of step-like variations of the mean permittivity at the grating boundaries that this type of resonant coupling can occur. A new method for the analysis of EAS and GAS in anisotropic gratings is developed. This method is based on the consideration of the diffractional divergence of the scattered wave and the two-wave approximation in anisotropic gratings. Special efforts are focused on the analysis of EAS and GAS of extraordinary waves in uniaxial gratings. In particular, it is demonstrated that increasing curvature of the normal surface in the direction of propagation of the scattered wave results in increase of its diffraction divergence and the resonant amplitude. A theoretical model is developed for comparison of the theoretical predictions with data obtained from experimental observations of EAS in a holographic grating written in a photorefractive medium. The developed model is applied for the interpretation of experimental observations of EAS in BaTiO3 photorefractive crystals. Good agreement with the theoretical predictions is demonstrated.

Submitted by Fabio Sobreira Campos da Costa (fabio.sobreira@ufpe.br) on 2016-07-11T12:33:28Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) silas-dissertacao-versao-final-2016.pdf: 1708159 bytes, checksum: 6c0efc5f2f0b27c79306418c9de516f1 (MD5) / Made available in DSpace on 2016-07-11T12:33:28Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) silas-dissertacao-versao-final-2016.pdf: 1708159 bytes, checksum: 6c0efc5f2f0b27c79306418c9de516f1 (MD5) Previous issue date: 2015-02-27 / FACEPE / A extração de conhecimento em ambientes com fluxo contínuo de dados é uma atividade que vem crescendo progressivamente. Diversas são as situações que necessitam desse mecanismo, como o monitoramento do histórico de compras de clientes; a detecção de presença por meio de sensores; ou o monitoramento da temperatura da água. Desta maneira, os algoritmos utilizados para esse fim devem ser atualizados constantemente, buscando adaptar-se às novas instâncias e levando em consideração as restrições computacionais. Quando se trabalha em ambientes com fluxo contínuo de dados, em geral não é recomendável supor que sua distribuição permanecerá estacionária. Diversas mudanças podem ocorrer ao longo do tempo, desencadeando uma situação geralmente conhecida como mudança de conceito (concept drift). Neste trabalho foi realizado um estudo comparativo entre alguns dos principais métodos de detecção de mudanças: ADWIN, DDM, DOF, ECDD, EDDM, PL e STEPD. Para execução dos experimentos foram utilizadas bases artificiais – simulando mudanças abruptas, graduais rápidas, e graduais lentas – e também bases com problemas reais. Os resultados foram analisados baseando-se na precisão, tempo de execução, uso de memória, tempo médio de detecção das mudanças, e quantidade de falsos positivos e negativos. Já os parâmetros dos métodos foram definidos utilizando uma versão adaptada de um algoritmo genético. De acordo com os resultados do teste de Friedman juntamente com Nemenyi, em termos de precisão, DDM se mostrou o método mais eficiente com as bases utilizadas, sendo estatisticamente superior ao DOF e ECDD. Já EDDM foi o método mais rápido e também o mais econômico no uso da memória, sendo superior ao DOF, ECDD, PL e STEPD, em ambos os casos. Conclui-se então que métodos mais sensíveis às detecções de mudanças, e consequentemente mais propensos a alarmes falsos, obtêm melhores resultados quando comparados a métodos menos sensíveis e menos suscetíveis a alarmes falsos. / Knowledge extraction from data streams is an activity that has been progressively receiving an increased demand. Examples of such applications include monitoring purchase history of customers, movement data from sensors, or water temperatures. Thus, algorithms used for this purpose must be constantly updated, trying to adapt to new instances and taking into account computational constraints. When working in environments with a continuous flow of data, there is no guarantee that the distribution of the data will remain stationary. On the contrary, several changes may occur over time, triggering situations commonly known as concept drift. In this work we present a comparative study of some of the main drift detection methods: ADWIN, DDM, DOF, ECDD, EDDM, PL and STEPD. For the execution of the experiments, artificial datasets were used – simulating abrupt, fast gradual, and slow gradual changes – and also datasets with real problems. The results were analyzed based on the accuracy, runtime, memory usage, average time to change detection, and number of false positives and negatives. The parameters of methods were defined using an adapted version of a genetic algorithm. According to the Friedman test with Nemenyi results, in terms of accuracy, DDM was the most efficient method with the datasets used, and statistically superior to DOF and ECDD. EDDM was the fastest method and also the most economical in memory usage, being statistically superior to DOF, ECDD, PL and STEPD, in both cases. It was concluded that more sensitive change detection methods, and therefore more prone to false alarms, achieve better results when compared to less sensitive and less susceptible to false alarms methods.

La photolithographie optique assure en partie à la microélectronique la miniaturisation des circuits électroniques. Afin de faire face à la limite de résolution de l'équipement de photolithographie, les industriels ont mis au point des techniques d'amélioration de la résolution dont certaines sont basées sur l'utilisation de la modélisation numérique. Jusqu'au nœud technologique 45 nm, cette modélisation ne prenait pas en compte la présence de plusieurs empilements de matériaux sous la résine photosensible négligeant ainsi les phénomènes de réflexion, de diffraction et d'ondes stationnaires. Pour les nœuds 32 nm et suivants, ces phénomènes rendent difficile le contrôle de la forme et des dimensions des motifs résine notamment pour les niveaux dont l'exposition s'effectue sans couche antireflet. Cette thèse CIFRE entre le laboratoire Hubert Curien de Saint- Etienne et l'industriel STMicroelectronique de Crolles traite de la caractérisation, de la modélisation et de la simulation numérique des effets d'empilement sous la résine photosensible. Le premier chapitre regroupe un ensemble de pensées sur la microélectronique, son histoire et définit les notions essentielles de ce domaine et de la modélisation numérique. Les chapitres deux et trois donnent respectivement l'état de l'art de la photolithographie optique et des techniques de correction des effets de proximité optique. Le chapitre quatre présente l'étude expérimentale, de la conception des structures test à la caractérisation des effets d'empilement en passant par le protocole de création des groupes de données. La prise en compte de ces effets est l'objet du chapitre cinq avec l'état de l'art des techniques existantes suivi de la description de l'algorithme de construction de modèles développé dans cette thèse. Enfin l'application de la méthode des sources généralisées à la photolithographie optique est évaluée dans le chapitre six / In IC manufacturing, optical photolithography is one of key actors of electronic circuit miniaturization. To work around the photolithography resolution limit, manufacturers developed resolution improvement techniques, including some based on numerical modeling. For nodes larger than 45 nm, this modeling didn't take into account several stacks under the photoresist and that caused optical reflection, diffraction, and standing wave phenomena to be neglected. For 32 nm and smaller nodes, these phenomena make it di cult to control the shape and dimensions of resist patterns, especially for layers without an anti-reflecting coating during exposure. The CIFRE thesis from Hubert Curien Laboratories in Saint-Etienne and industrial STMicroelectronics from Crolles deals with wafer stack effect characterization, modeling, and numerical simulation. The first chapter gives my philosophy and history of IC manufacturing, and defines concepts in this field and concepts about numerical modeling. Chapter Two discusses state-of-the-art optical photolithography and Chapter Three discusses state-of-the-art optical proximity correction. Chapter Four emphasizes an experimental study from test pattern conception to wafer stack effect characterization, including data set building protocol. Chapter Five covers wafer stack effect management, first describing the current status of the industry followed by a description of the model building algorithm developed during this thesis. Finally, Chapter Six assesses the generalized source method applied to the photolithography process simulation

This thesis fills in the gaps in the existing theory of wave phenomena in thick diffraction gratings at extreme angles of scattering, i.e. when the scattered wave propagates parallel or almost parallel to the grating boundaries. A consistent theory of a new type of Bragg scattering of bulk and guided optical modes in thick uniform and non-uniform, dissipative and non-dissipative, slanted periodic gratings has been developed. This type of scattering is called extremely asymmetrical scattering (EAS). One of the main distinctive features of EAS is the strong resonant increase of the scattered wave amplitude compared to the amplitude of the incident wave. Several unique combinations of strong resonances shaping a complex multi-resonant pattern of EAS in different types of gratings have been predicted and investigated theoretically and numerically. This includes the prediction of a new resonant wave effect in non-uniform gratings with varying phase – double-resonant EAS, the discovery of several sharp and strong resonances with respect to scattering angle in gratings with the scattered wave propagating almost parallel to the grating boundaries (grazing-angle scattering (GAS)) for the case of second-order scattering, and the prediction of a new type of eigenmode in gratings with second-order scattering (especially in gratings with large amplitude). In addition, several other important practical problems that may be crucial for the experimental observation and application of EAS and GAS have been solved. These are the determination of the tolerance of EAS to small grating imperfections, e.g., fluctuations of the grating amplitude, prediction of unusually high sensitivity of second-order EAS to small variations of mean structural parameters, determination of the effect of weak dissipation on EAS, etc. Physical reasons for the predicted resonances and effects are explained. In particular, the crucial role of the diffractional divergence for EAS and GAS has been revealed, especially for non-uniform gratings. Methods of analysis involve the approximate and rigorous approaches. The approximate method is based on understanding the role of the diffractional divergence in the geometry of EAS and the two-wave approximation (valid for any types of waves). The rigorous approach is based on the rigorous coupled-wave analysis (RCWA) and, in particular, the known enhanced T-matrix algorithm (by Moharam, et al.) that is numerically stable for narrow and wide gratings with arbitrary amplitude (valid only for bulk electromagnetic waves).

Tout d'abord, des ondes de surface, solutions de problèmes aux limites hyperboliques non linéaires, sont étudiées : on construit une solution BKW sous forme de développement infini en puissance de epsilon. On le justifie rigoureusement, en construisant une solution exacte, qui admet ce développement asymptotique. On montre que la solution n'est pas nécessairement purement localisée sur la frontière, même lorsque le terme source l'est ; l'exemple d'un cas particulier de l'élasticité est traité. Ensuite, on étudie la réflexion d'ondes non linéaires discontinues, pour des problèmes aux limites hyperboliques, faiblement bien posés, ni fortement stables, ni fortement instables. On étudie comment les singularités d'une solution striée sont réfléchies lorsque la solution atteint la frontière. On prouve des estimations striées et en normes infinies. On montre qu'une discontinuité du gradient de la solution à travers un hyperplan peut être réfléchie en une discontinuité de la solution elle-même. / We first study surface waves, solutions of hyperbolic nonlinear boundary value problems. We construct BKW solutions in the weakly nonlinear regime with infinite expansion in powers of ε. We rigorously justify this expansion,constructing exact solutions, which admit the asymptotic expansions. We also show that the solution is not necessarily localized at the order O(ε∞) in the interior, even if the data are ; a particular case of elasticity is studied: we prove that fast oscillatory elastic surface waves can produce non trivial internal non oscillatory displacements.Afterwards, we study the reflection of non linear discontinuous waves, for weakly well-posed hyperbolic boundary value problems, satisfying the (WR) condition, which has been introduced in [1, 12], that is in a case where the IBVP is neither strongly stable, nor strongly unstable. We study how the singularities of a striated solution are reflected when the solution hits the boundary. We prove striated estimates and L∞ estimates and observe the loss of one derivative: we show that a discontinuityof the gradient of the solution across an hyperplane can be reflected in a discontinuity across an hyperplane of the solution itself.

L'implantation en Virgule Flottante (VF) d'une fonction à valeurs réelles est réalisée avec arrondi correct si le résultat calculé est toujours égal à l'arrondi de la valeur exacte, ce qui présente de nombreux avantages. Mais pour implanter une fonction avec arrondi correct de manière fiable et efficace, il faut résoudre le «dilemme du fabricant de tables» (TMD en anglais). Deux algorithmes sophistiqués (L et SLZ) ont été conçus pour résoudre ce problème, via des calculs longs et complexes effectués par des implantations largement optimisées. D'où la motivation d'apporter des garanties fortes sur le résultat de ces pré-calculs coûteux. Dans ce but, nous utilisons l'assistant de preuves Coq. Tout d'abord nous développons une bibliothèque d'«approximation polynomiale rigoureuse», permettant de calculer un polynôme d'approximation et un intervalle bornant l'erreur d'approximation à l'intérieur de Coq. Cette formalisation est un élément clé pour valider la première étape de SLZ, ainsi que l'implantation d'une fonction mathématique en général (avec ou sans arrondi correct). Puis nous avons implanté en Coq, formellement prouvé et rendu effectif 3 vérifieurs de certificats, dont la preuve de correction dérive du lemme de Hensel que nous avons formalisé dans les cas univarié et bivarié. En particulier, notre «vérifieur ISValP» est un composant clé pour la certification formelle des résultats générés par SLZ. Ensuite, nous nous sommes intéressés à la preuve mathématique d'algorithmes VF en «précision augmentée» pour la racine carré et la norme euclidienne en 2D. Nous donnons des bornes inférieures fines sur la plus petite distance non nulle entre sqrt(x²+y²) et un midpoint, permettant de résoudre le TMD pour cette fonction bivariée. Enfin, lorsque différentes précisions VF sont disponibles, peut survenir le phénomène de «double-arrondi», qui peut changer le comportement de petits algorithmes usuels en arithmétique. Nous avons prouvé en Coq un ensemble de théorèmes décrivant le comportement de Fast2Sum avec double-arrondis. / The Floating-Point (FP) implementation of a real-valued function is performed with correct rounding if the output is always equal to the rounding of the exact value, which has many advantages. But for implementing a function with correct rounding in a reliable and efficient manner, one has to solve the ``Table Maker's Dilemma'' (TMD). Two sophisticated algorithms (L and SLZ) have been designed to solve this problem, relying on some long and complex calculations that are performed by some heavily-optimized implementations. Hence the motivation to provide strong guarantees on these costly pre-computations. To this end, we use the Coq proof assistant. First, we develop a library of ``Rigorous Polynomial Approximation'', allowing one to compute an approximation polynomial and an interval that bounds the approximation error in Coq. This formalization is a key building block for verifying the first step of SLZ, as well as the implementation of a mathematical function in general (with or without correct rounding). Then we have implemented, formally verified and made effective 3 interrelated certificates checkers in Coq, whose correctness proof derives from Hensel's lemma that we have formalized for both univariate and bivariate cases. In particular, our ``ISValP verifier'' is a key component for formally verifying the results generated by SLZ. Then, we have focused on the mathematical proof of ``augmented-precision'' FP algorithms for the square root and the Euclidean 2D norm. We give some tight lower bounds on the minimum non-zero distance between sqrt(x²+y²) and a midpoint, allowing one to solve the TMD for this bivariate function. Finally, the ``double-rounding'' phenomenon can typically occur when several FP precision are available, and may change the behavior of some usual small FP algorithms. We have formally verified in Coq a set of results describing the behavior of the Fast2Sum algorithm with double-roundings.