Thesis (doctoral)--Rheinische Friedrich-Wilhelms-Universität, Bonn, 2005. / Includes bibliographical references (p. 206-213).

Texte rem. de: Thesis (doct.)--Frankfurt-am-Main--Universität, 1985. / Notes bibliogr. Bibliogr. p. 205-213.

L'analyse pertinente de l'information contenue dans les mesures instrumentales modernes nécessite le développement de méthodes algorithmiques performantes. La spectroscopie n'échappant pas à cette règle, nombre d'outils chimiométriques d'analyse multivariée ont été mis au point en conséquence. L'objectif de cette thèse est de participer à ces développements pour caractériser les spectres et les profils de concentration des composés intervenant lors de réactions photoinduites suivies en spectroscopique résolue en temps. Un premier travail concerne l'analyse du rang de données de différence. Une déficience de rang systématique de ces données a été mise en évidence. Une approche basée sur un algorithme de modélisation hybride est proposée en réponse à ce problème. En effet l'introduction d'un modèle cinétique compense cette déficience de rang des données. Le deuxième travail se rapporte à la caractérisation de l'influence du bruit (d'une part sur la mesure et d'autre part sur la résolution multivariée). Une analyse de la structure de corrélation du bruit permet d'expliquer les phénomènes de surajustement des méthodes de modélisation souple. Un second algorithme résout ce problème en pondérant les données en fonction d'une estimation de l'erreur de mesure. Ces points sont illustrés par deux applications; premièrement une contribution à l'analyse du centre réactionnel de la bactérie Rhodobacter sphaeroides par spectroscopie lRTF et ensuite l'étude de la photorelaxation de la benzophénone par spectroscopie d'absorption transitoire. / Pertinent analysis of the information embedded in modern instrumental measurements requires the development of efficient mathematical and statistical tools. ln spectroscopy, various tools have been proposed in such purpose. The aim of this thesis is to participate to these developments in order to characterize the spectra and concentration profiles of the compounds involved in photoinduced reactions monitored by time-resolved spectroscopy. A first part of this work is related to the rank analysis of difference data. A systematic rank deficiency has been demonstrated in this case. An approach based on a hybrid algorithm is proposed as a solution to this problem. lndeed the introduction of a kinetic model compensates this rank deficiency. A second part of the work deals with the characterisation of the noise influence (firstly on measurement and then on multivariate resolution). An analysis of the correlations structure of the noise permit to explain the overfitting phenomenon observed with soft modelling methods. A second algorithm solves this problem through a weighting scheme involving an estimation of the measurement error.These points are illustrated by two applications: first, a contribution to the analysis of the reaction centre of Rhodobacter sphaeroides by FTIR spectroscopy; then the study of the benzophenone photorelaxation by transient absorption spectroscopy.

University, Diss., 2001--München.

NOGUEIRA, Vanessa Lúcia Rodrigues. Papel de proteínas na tinta liberada pela lesma-do-mar Aplysia dactylomela Rang 1828 os mecanismos de defesa do animal. 2005. 90 f. Dissertação (Mestrado em Ciências Marinhas Tropicais) - Instituto de Ciências do Mar, Universidade Federal do Ceará, Fortaleza, 2005. / Submitted by Debora Oliveira (deby_borboletinha@hotmail.com) on 2011-12-22T15:10:49Z No. of bitstreams: 1 2005_dis_vlrnogueira.pdf: 12654022 bytes, checksum: 35936f35caac0bc23dbf25e1b2ffcafc (MD5) / Approved for entry into archive by Nadsa Cid(nadsa@ufc.br) on 2012-01-23T17:07:24Z (GMT) No. of bitstreams: 1 2005_dis_vlrnogueira.pdf: 12654022 bytes, checksum: 35936f35caac0bc23dbf25e1b2ffcafc (MD5) / Made available in DSpace on 2012-01-23T17:07:24Z (GMT). No. of bitstreams: 1 2005_dis_vlrnogueira.pdf: 12654022 bytes, checksum: 35936f35caac0bc23dbf25e1b2ffcafc (MD5) Previous issue date: 2005 / The sea hare Aplysia dactylomela is known by discharging a purple ink when disturbed. As it doesn't show any external structure of protection, it is believed that this secretion, rich in bioactive substances, participates in the chemical defense against pathogens and/or predators. The ink is constituted mainly by pigments, proteins and low molecular mass substances. The pigments are originated from red algae, nevertheless, there is lack of information about the origin, processing, storage and function of the ink proteins. This work describes the protein composition of the ink and present some physiochemical and biological characteristics of the dactylomelin-P, an antibacterial protein from the ink besides information concerning the location of this protein in the purple gland. The ink was obtained from specimens collected at Fleixeiras beach, Ce. The protein composition of the ink was studied by bi-dimensional electrophoresis. The purification of dactylomelin-P consisted basically in ion exchange and hydrophobic interaction chromatography. Dactylomelina-P was analyzed as to molecular mass, isoeletric point, amino acid composition, carbohydrates, heat stability, pH and proteinase resistance. Several biological activities were tested with the ink and with dactylomelina-P, including antimicrobial, enzymatic, haemagglutinating, anticoagulant, hemolytic, cytotoxic and toxic activities. The immunolocalization assays of dactylomelina-P were carried out in different sea hare tissues by western blot and immunohistochemistry and the interaction with Staphylococcus aureus was investigated by immunocytochemistry. The ink of A. dactylomela contains more than 40 proteins/peptides, with molecular masses below 70 kDa and acid pIs. The most abundant protein in the ink is the dactylomelin-P, a monomeric protein with 59,8 kDa, pI 5,0, high methionin content, and less than 1% carbohydrates. It is denatured at 60 ºC for 10 minutes and it resists to pH range of 3-12. Dactylomelin-P shows a wide antibacterial action spectrum, but no antifungal, unlike the ink that possesses some factor with this activity. It is particularly efficient against sea bacteria, can be bactericidal (4,0μg/ml) or bacteriostatic (0,2μg/ml), depending on the concentration. Dactylomelin-P agglutinated rabbit, mice and rat erythrocytes, but it did not show anticoagulating, hemolytic and cytotoxic activities. The LD50 for mice was 60-100 mg/Kg, being considered moderately toxic. Dactylomelin-P was only found in the ink gland, being located preferentially in the cells of the producing vesicles. The electronic transmission microscopy revealed that the dactylomelina-P crosses the cell wall of S. aureus and interacts mainly with the plasmatic membrane, probably interfering in the metabolism of the bacterium, instead of causing mechanic damage to the cell. / O gastrópode marinho Aplysia dactylomela é conhecido por liberar uma tinta púrpura sempre que é importunado. Como não possui nenhuma estrutura externa de proteção, acredita-se que essa secreção, rica em substâncias biologicamente ativas, participe da defesa química do animal. A tinta é composta de pigmentos, proteínas e substâncias de baixa massa molecular. Os pigmentos da tinta são originados de algas vermelhas, mas quanto às proteínas, pouco é conhecido sobre a sua origem, processamento, local de armazenamento e função no animal. Este trabalho descreve pela primeira vez a composição protéica da tinta e apresenta algumas características físico-químicas e biológicas da dactylomelina-P, uma proteína antibacteriana presente na tinta dessa lesma, além de trazer informações acerca da localização desta proteína na glândula de tinta. A tinta foi obtida a partir de espécimes encontrados na praia de Fleixeiras, Ce. A composição de proteínas da tinta foi determinada por eletroforese bidimensional e a purificação da dactylomelina-P foi feita através de cromatografias de troca iônica e interação hidrofóbica. Dactylomelina-P foi analisada quanto à massa molecular, ponto isoelétrico, composição de aminoácidos, presença de carboidratos, estabilidade térmica, resistência ao pH e a proteases. Várias atividades biológicas foram testadas com a tinta e com a dactylomelina-P, incluindo atividades antimicrobianas, enzimáticas, hemaglutinante, anticoagulante, hemolítica, citotóxica e tóxica. Os ensaios de localização da proteína foram realizados em diferentes tecidos da lesma por western blot, na glândula de tinta por imunohistoquímica e a interação com a bactéria Staphylococcus aureus foi feita por imunocitoquímica. A tinta de A. dactylomela contém mais de 40 proteínas/ peptídeos, com massas abaixo de 70 kDa e pIs na faixa ácida. A proteína mais abundante na tinta é a dactylomelina-P, que é uma molécula monomérica, de 59,8 kDa, pI 5,0, que possui alto teor de metionina e menos de 1% de carboidratos, é desnaturada a 60 ºC por 10 minutos e resiste a pHs entre 3-12. Dactylomelina-P mostra um amplo espectro de ação antibacteriano, mas não antifúngico, ao contrário da tinta que possui um fator com esta atividade. É particularmente eficiente contra bactérias marinhas, podendo ser bactericida (4,0μg/ml) ou bacteriostática (0,2μg/ml), dependendo da concentração. Dactylomelina-P aglutinou eritrócitos de coelho, ratos e camundongos, não apresentou atividade anticoagulante, hemolítica e nem citotóxica. A DL50 para camundongos ficou entre 60-100 mg/Kg, sendo considerada moderadamente tóxica. Dactylomelina-P só foi encontrada na glândula de tinta, localizando-se preferencialmente nas células das vesículas produtoras. Ensaios de microscopia eletrônica de transmissão revelaram que a dactylomelina atravessa a parede celular da bactéria S. aureus e, interage principalmente com a membrana citoplasmática, provavelmente interferindo no metabolismo, ao invés de causar danos à célula.

Le développement de méthodes de décomposition de tableaux multi-dimensionnels suscite toujours autant d'attention, notamment d'un point de vue applicatif. La plupart des algorithmes, de décompositions tensorielles, existants requièrent une estimation du rang du tenseur et sont sensibles à une surestimation de ce dernier. Toutefois, une telle estimation peut être difficile par exemple pour des rapports signal à bruit faibles. D'un autre côté, estimer simultanément le rang et les matrices de facteurs du tenseur ou du tenseur cœur n'est pas tâche facile tant les problèmes de minimisation de rang sont généralement NP-difficiles. Plusieurs travaux existants proposent d'utiliser la norme nucléaire afin de servir d'enveloppe convexe de la fonction de rang. Cependant, la minimisation de la norme nucléaire engendre généralement un coût de calcul prohibitif pour l'analyse de données de grande taille. Dans cette thèse, nous nous sommes donc intéressés à l'approximation d'un tenseur bruité par un tenseur de rang faible. Plus précisément, nous avons étudié trois modèles de décomposition tensorielle, le modèle CPD (Canonical Polyadic Decomposition), le modèle BTD (Block Term Decomposition) et le modèle MTD (Multilinear Tensor Decomposition). Pour chacun de ces modèles, nous avons proposé une nouvelle méthode d'estimation de rang utilisant une métrique moins coûteuse exploitant la parcimonie de groupe. Ces méthodes de décomposition comportent toutes deux étapes : une étape d'estimation de rang, et une étape d'estimation des matrices de facteurs exploitant le rang estimé. Des simulations sur données simulées et sur données réelles montrent que nos méthodes présentent toutes une plus grande robustesse à la présence de bruit que les approches classiques. / Last decades, tensor decompositions have gained in popularity in several application domains. Most of the existing tensor decomposition methods require an estimating of the tensor rank in a preprocessing step to guarantee an outstanding decomposition results. Unfortunately, learning the exact rank of the tensor can be difficult in some particular cases, such as for low signal to noise ratio values. The objective of this thesis is to compute the best low-rank tensor approximation by a joint estimation of the rank and the loading matrices from the noisy tensor. Based on the low-rank property and an over estimation of the loading matrices or the core tensor, this joint estimation problem is solved by promoting group sparsity of over-estimated loading matrices and/or the core tensor. More particularly, three new methods are proposed to achieve efficient low rank estimation for three different tensors decomposition models, namely Canonical Polyadic Decomposition (CPD), Block Term Decomposition (BTD) and Multilinear Tensor Decomposition (MTD). All the proposed methods consist of two steps: the first step is designed to estimate the rank, and the second step uses the estimated rank to compute accurately the loading matrices. Numerical simulations with noisy tensor and results on real data the show effectiveness of the proposed methods compared to the state-of-the-art methods.

Dans beaucoup de cas les nuisances sonores auxquelles nous sommes quotidiennement exposés sont dues à la vibration d'une structure (machine industrielle, véhicule, appareil ménager …). Néanmoins, tous les bruits que nous percevons ne sont pas forcément dûs à la vibration d'une structure, par exemple, les bruits aérodynamiques, les bruits de turbine ou les bruits de jet … La recherche en vibroacoustique est étroitement liée avec des applications industrielles, car l'industrie a besoin des nouveaux outils numériques, développés dans les centres de recherche, pour concevoir de nouveaux produits silencieux. En effet, les démarches purement expérimentales sont en général longues, compliquées et coûteuses, elles peuvent être, de plus, très peu efficaces. Puisque l'objectif ultime est la conception d'une structure qui permet de réduire le bruit pour un très bon confort acoustique, les simulations numériques peuvent être incluses dans l'optimisation de la conception avec des techniques de conception optimales de forme et l'optimisation des matériaux. Une fois les simulations validées par les résultats expérimentaux, elles peuvent être utilisées comme outil de conception pour l'amélioration de la structure du système concerné. L'objectif principal de mon travail de thèse est le développement des outils de prédiction numériques permettant la réduction des nuisances sonores dues à la vibration des structures. Pour ce faire, des formulations théoriques originales ont été formulées, puis implantées afin de favoriser la conception de produits silencieux. D'une manière plus spécifique, deux parties vont être traitées : La première partie aborde le problème bien connu des fréquences irrégulières de la méthode des éléments finis de frontière, la BEM, pour le rayonnement acoustique dans un domaine extérieur. Dans la deuxième partie de cette thèse la formulation de la méthode multipôlaire rapide FMM couplée à la BEM, ainsi que sa mise en œuvre et validation ont été effectuées afin de repousser les limites de la BEM en terme de temps de calcul ainsi que de mémoire. / In many cases the noise which we are daily exposed are due to the vibration of a structure (industrial machinery, vehicle, appliance...). Nevertheless, all the sounds we perceive are not necessarily due to the vibration of a structure, for example, wind noise, the sounds of turbine or jet noise...The vibroacoustic research is closely linked with industrial applications because the industry needs new numerical tools, developed in research centers to develop new silent products. Indeed, purely experimental approaches are generally lengthy, complicated and expensive they can be, again, very inefficient. Since the ultimate objective is to design a structure that reduces noise for a good acoustic comfort, numerical simulations can be included in the design optimization techniques to design optimum shape and optimizationmaterials. Once the simulations validated by experimental results, it can be used as a design tool for improving the structure of the affected system. The main aim of my thesis is the development of numerical predictive tools for the noise reduction due to the vibrationof structures. To do this, the original theoretical formulations have been developed and implemented to encourage the design of silent products. In a more specific way, both parties will be addressed : the first part addreeses the familiarproblem of irregular frequencies of the finite element boundary, the BEM for acoustic radiation in an external field. In the second part of this thesis the formulation of the fast multipole method FMM coupled with BEM, as well asits implementation and validation were carried out to push the boundaries of the BEM in terms of computation time and memory.

Texte remanié de: Th. Ph. D. / Bibliogr. p. 213-227.

En introduktion ges till portföljteori och tillämpning av tangentportföljen med en riskfri tillgång presenteras. Nyttofunktioner förklaras och hur tangentportföljen kan härledas med hjälp av en nyttofunktion. Annan bakgrundsinformation som är viktig för att förstå resultat från [1] om vikter i tangentportföljen då kovariansmatrisen inte är inverterbar presenteras också. Teorin tillämpas på historisk data från S&P. Resultaten visas i grafer och tabeller som analyseras.

Common cameras with a dynamic range of two orders cannot reproduce typical outdoor scenes with a radiance range of over five orders. Most high dynamic range (HDR) imaging techniques reconstruct the whole dynamic range from exposure bracketed low dynamic range (LDR) images. But the camera must be kept steady with no or small motion, which is not practical in many cases. Thus, we develop a more efficient framework for omnidirectional HDR imaging with a moving camera. The proposed framework is composed of three major stages: geometric calibration and rotational alignment, multi-view stereo correspondence and HDR composition. First, camera poses are determined and omnidirectional images are rotationally aligned. Second, the aligned images are fed into a spherical vision toolkit to find disparity maps. Third, enhanced disparity maps are used to warp differently exposed neighboring images to a target view and an HDR radiance map is obtained by fusing the registered images in radiance. We develop disparity-based forward and backward image warping algorithms for spherical stereo vision and implement them in GPU. We also explore some techniques for disparity map enhancement including a superpixel technique and a color model for outdoor scenes. We examine different factors such as exposure increment step size, sequence ordering, and the baseline between views. We demonstrate the success with indoor and outdoor scenes and compare our results with two state-of-the-art HDR imaging methods. The proposed HDR framework allows us to capture HDR radiance maps, disparity maps and an omnidirectional field of view, which has many applications such as HDR view synthesis and virtual navigation.