La thèse s’inscrit dans un contexte d’essor de l’audio spatialisé (5.1, Dolby Atmos...). Parmi les formats audio 3D existants, l’ambisonie permet une représentation spatiale homogène du champ sonore et se prête naturellement à des manipulations : rotations, distorsion du champ sonore. L’objectif de cette thèse est de fournir un outil d’analyse et de manipulation de contenus audio (essentiellement vocaux) au format ambisonique. Un fonctionnement temps-réel et en conditions acoustiques réelles sont les principales contraintes à respecter. L’algorithme mis au point est basé sur une analyse en composantes indépendantes (ACI) appliquée trame à trame qui permet de décomposer le champ acoustique en un ensemble de contributions, correspondant à des sources (champ direct) ou à de la réverbération. Une étape de classification bayésienne, appliquée aux composantes extraites, permet alors l’identification et le dénombrement des sources sonores contenues dans le mélange. Les sources identifiées sont localisées grâce à la matrice de mélange obtenue par ACI, pour fournir une cartographie de la scène sonore. Une étude exhaustive des performances est menée sur des contenus réels en fonction de plusieurs paramètres : nombre de sources, environnement acoustique, longueur des trames, ou ordre ambisonique utilisé. Des résultats fiables en terme de localisation et de comptage de sources ont été obtenus pour des trames de quelques centaines de ms. L’algorithme, exploité comme prétraitement dans un prototype d’assistant vocal domestique, permet d’améliorer significativement les performances de reconnaissance, notamment en prise de son lointaine et en présence de sources interférentes. / The context of this thesis is the development of spatialized audio (5.1 contents, Dolby Atmos...) and particularly of 3D audio. Among the existing 3D audio formats, Ambisonics and Higher Order Ambisonics (HOA) allow a homogeneous spatial representation of a sound field and allows basics manipulations, like rotations or distorsions. The aim of the thesis is to provides efficient tools for ambisonics and HOA sound scene analyse and manipulations. A real-time implementation and robustness to reverberation are the main constraints to deal with. The implemented algorithm is based on a frame-by-frame Independent Component Analysis (ICA), wich decomposes the sound field into a set of acoustic contributions. Then a bayesian classification step is applied to the extracted components to identify the real sources and the residual reverberation. Direction of arrival of the sources are extracted from the mixing matrix estimated by ICA, according to the ambisonic formalism, and a real-time cartography of the sound scene is obtained. Performances have been evaluated in different acoustic environnements to assess the influence of several parameters such as the ambisonic order, the frame length or the number of sources. Accurate results in terms of source localization and source counting have been obtained for frame lengths of a few hundred milliseconds. The algorithm is exploited as a pre-processing step for a speech recognition prototype and allows a significant increasing of the recognition results, in far field conditions and in the presence of noise and interferent sources.

La préservation de la richesse aquatique est devenue un souci majeur d'ordre mondial suite au risque de pénurie en eau. Au Liban, les rejets anthropiques incontrôlés et incontrôlables au bord des rivières menacent la qualité de ses eaux de surface. Pour cela, quatre cours d'eau libanais ont été choisis comme cadre d'étude : la rivière Damour, la rivière Kadisha-Abou Ali, la rivière Ibrahim et la rivière Oronte.Les différents paramètres physico-chimiques et microbiologiques étudiés ont permis dans un premier temps la construction d'une base de données de chacune de ces rivières. Une seule analyse spatio-temporelle des paramètres séparément n'aide pas à définir l'état trophique des rivières. Partant de l'idée des corrélations que peuvent exister entre certains paramètres, l'Analyse en Composante Principale (ACP) retenant la totalité de l'information sera utilisée en dépit des méthodes classiques. Cet outil statistique a permis de classer les rivières par niveau de pollution. Il a aussi aidé à observer l'impact des apports des polluants sur les différentes stations de rivières étudiées.Pour suivre le devenir des coliformes fécaux dans les eaux, l'ACP des variables microbiologiques a montré la persistance des colonies bactériennes dans les eaux malgré les conditions climatiques diverses et le régime méditerranéen torrentiel des rivières. En effet, les sédiments constituent des réservoirs potentiels de microorganismes pathogènes. Le décrochage bactérien des agrégats sédimentaires et la remise en suspension dans l'eau pose un problème alarmant de santé publique. / Preservation of aquatic wealth has worldwide become a major concern due to the risk of water shortage. In Lebanon, uncontrolled and uncontrollable anthropogenic rejections along rivers threaten the quality of its surface waters. Four Lebanese rivers were selected as study framework: The Damour river, Kadisha-Abu Ali river, Ibrahim river and Orontes river. The physico-chemical and microbiological parameters analyzed has formed a large database of these rivers. The spatio-temporal analysis of separate parameters did not help defining the trophic status of rivers. The assumption that correlations exist between certain parameters, guides us to use the Principal Component Analysis (PCA) in spite of conventional methods. This statistical tool was used to define the pollution levels in rivers. It also leads to observe the impact of pollutants inflows on different sites studied. To follow the fate of fecal coliform in water, the ACP of microbiological variables showed the persistence of bacterial colonies in water despite the various climatic conditions and the Mediterranean flow rate. Indeed, sediments are potential reservoirs of pathogenic microorganisms. The bacteria aggregated to the sediment can be present again in water column causing an alarming public health problem.

L’objectif de ce travail de thèse est d’envisager les apports de l’Emission Acoustique (EA) en tant qu’outil de diagnostic et de surveillance des structures et ouvrages en bois. Nous avons appliqué cet outil et développé des méthodes de suivi de la fissuration. Si l’objectif affiché est l’utilisation de l’EA pour de la surveillance in-situ, un passage par des manipulations en laboratoire a été nécessaire pour valider les méthodes mises en œuvre. Des essais de chargement à déplacement imposé ont été réalisés sur des éprouvettes de Douglas (DCB à inertie constante) soumises à un chargement en mode I, pour deux teneurs en eaux distinctes. En parallèle du système d’acquisition acoustique, un système d’acquisition d’images a été utilisé pour suivre l’évolution de la fissuration sur les faces principales des éprouvettes.Dans un premier temps, nous avons constaté que la cinétique de fissuration était corrélée à l’énergie acoustique des signaux enregistrés. Une écoute globale, sans filtration des données, fournit une indication sur l’évolution de l’endommagement d’une structure.En effectuant un premier traitement des données, en considérant les évènements acoustiques et en utilisant des courbes de correction de localisation et d’amplitude des signaux, nous avons pu déterminer la position d’un (ou plusieurs) pic(s) d’activité acoustique. La courbe d’évolution de la position de ce(s) pic(s) d’activité donne un résultat similaire à celle de la pointe de fissure obtenue avec le système d’imagerie. Nous avons ainsi pu estimer un taux de restitution d’énergie moyen Gc. Dans le but de passer à des applications in situ, nous avons mis en œuvre cette méthode lors d’un essai de fluage sur une éprouvette à inertie variable. L’EA a su prendre le relais lorsque les relevés visuels étaient absents. L’EA a aussi pu fournir des informations complémentaires sur l’initiation et la propagation de la fissuration et ce, sur une durée surveillance longue.La méthode de détection des pics d’activités par EA montre que les évènements acoustiques de forte amplitude sont localisés de part et d’autre de la pointe de fissure obtenue par imagerie. Dans un deuxième temps, en effectuant un parallèle avec les méthodes développées dans les matériaux composites, le bois pouvant y être assimilé par certains aspects, nous avons pu, par une utilisation conjointe des outils statistiques, de l’imagerie électronique (MEB) et de l’analyse des formes d’ondes, identifier les mécanismes de ruine présents lors des essais.Après un travail de filtration des données obtenues sur des essais à mode de ruine choisi (traction, flexion et cisaillement), quatre familles d’évènements ont pu être mis en évidence. Par l’étude de leurs formes d’ondes respectives, ces familles ont pu être associées à de la rupture de fibre, de la fissuration de matrice, de la décohésion et du délaminage telles qu’on les retrouve dans les matériaux composites. L’étude au MEB a confirmé qu’une multitude de mécanismes étaient présents sur les plans de fissuration des éprouvettes et la vraisemblance des résultats obtenus.D’un point de vue quantitatif, deux mécanismes de ruine principaux ressortent : la rupture de fibres et la fissuration matricielle. Il s’agit là d’indicateurs dont l’évolution peut constituer un indicateur, précurseur à la ruine d’un ouvrage. Le travail effectué dans cette thèse constitue un pas dans l’emploi de l’EA dans la surveillance des structures et des ouvrages en bois et laisse entrevoir d’autres emplois notamment en mécanique de la rupture et dans l’étude des zones d’élaboration. / The aim of this thesis is to consider what Acoustic Emission (AE) can provide as a tool for diagnosis and assessment of wood structures and works. We applied this tool and developed methods for tracking crack propagation. If our intention is using AE for the in-situ monitoring, we have to do experiments in laboratory to validate the methods developped. Loading tests under imposed displacement were conducted on Douglas samples (DCB constant inertia) subjected to a loading in mode I, for two levels of moisture content. In tandem of the acoustic emission emission system, an image acquisition system was used to record crack propagation on the main faces of the samples.First, we showed that the cracking kinetics correlates with the acoustic energy of the recorded signals. A global analysis, with unfiltered data, provides a good indication of the evolution of the damage within a structure. Then by performing a first data processing, by taking into account the acoustic events and by using correction curves of localization and amplitude, we were able to determine the position of one (or more) peak(s) of acoustic activity. The evolution curve of the position of this (these) peak(s) of activity gives a result similar of the crack tip evolution obtained with the digital image acquisition system. By this way we were able to estimate an average restoration level of critical energy Gc. In the purpose of making in situ applications, we have implemented this method during a creep test on a specimen with variable inertia. A previously loaded sample was placed to a relative humidity variation - from a humid atmosphere to a dry atmosphere. Only spot measurements of the crack tip position has been made. AE was able to take over when visual statements were absent. AE was also able to provide additional information on the initiation and propagation of cracks over a long term monitoring. The method for detecting peaks of activities by AE shows that acoustic events with high amplitude are located on both side of the crack tip obtained by imaging.Secondly, by comparing with the methods developed in composite materials, wood could be considered as such, we were able, by joint use of statistical tools, electronic imaging (SEM) and analysis of waveforms, to identify failure mechanisms which were present during testing. After processing data obtained on failure specific mode test (tensile, bending and shear), four clusters of events have been highlighted. By studying their respective waveforms and signal caracteristics, these families have been associated with the rupture of fiber, matrix cracking, debonding and delamination such as those found in composites. The study of SEM images made from owr samples has confirmed the presence of a multitude of mechanisms on the cracking path of the wood samples. This confirms the reasonableness of results obtained. From a quantitative perspective, two major failure mechanisms stand out: the breaking of fibers and matrix cracking. The evolution of these indicators may be a precursor to the ruin of a structure.The work undertaken in this thesis is a step in the use of AE in monitoring structures and wooden structures. It suggests other purposes for the use of AE in wood. We could imaging using AE in order to study the process zones by using mTDCB samples.

Les aires cérébrales entretiennent des relations fonctionnelles, formant ainsi des réseaux qui peuvent être altérés dans diverses pathologies. L'étude de ces réseaux de connectivité fonctionnelle pourrait potentiellement aider au diagnostic d'un individu et au traitement de sa pathologie. À travers quatre études, nous avons montré que l'analyse en composantes indépendantes spatiale est une méthode suffisamment sensible, reproductible et spécifique pour mettre en évidence, à l’échelle individuelle et au repos, des réseaux sains et pathologiques fiables. Ainsi, le réseau pathologique sous-tendant les hallucinations acoustico-verbales permet de définir les aires cérébrales à traiter par stimulation magnétique transcrânienne. Parmi les réseaux sains, ceux qui impliquent le cortex cingulaire postérieur et le précunéus semblent profondément altérés dans les troubles de la conscience, et peuvent servir d'outil diagnostic pour distinguer le locked-in syndrome de l'état végétatif. Il est désormais possible de cartographier, à l'échelle individuelle, les relations fonctionnelles entre les aires cérébrales. L’étude à venir de la dynamique et du niveau d’activité des réseaux de connectivité fonctionnelle nous renseignera davantage sur leurs fonctions et leur implication dans la pathologie. / Brain areas are functionally connected in networks, even at rest. Since such connectivity networks could be impaired in several pathologies, they could potentially serve for diagnosis and treatment. Based on four studies, spatial independent component analysis has shown sufficient sensitivity, reproducibility and specificity to produce reliable healthy as well as pathological networks at the individual level. Therefore, the network underlying auditory hallucination could define the brain areas to treat by transcranial magnetic stimulation. Among the common resting-state networks, the ones involving the posterior cingular cortex and the precuneus seem deeply altered in disorders of consciousness, and so could be used as a diagnostic tool to distinguish the locked-in syndrome from the vegetative state. We can now map, at the individual level, the functional relationships between brain areas. Further studies on the dynamic and on the level of activity of the functional connectivity networks might provide relevant information about their functions and their involvement in the pathology.

The main purpose of thesis is creation and evaluation of models for change detection of arable land to grassland by Hybrid-based Change Detection method, which combined approaches based on the Vegetation Indices, Image Differencing and Principal Component Analysis. Six locations with different seasonal configuration of images with high resolution and one locality covered by image with very high resolution were used. The areas were spread across the foothill areas of the Czech Republic. The selection of predictors and the most suitable model was supported by statistical calculation. Application selected models were carried out using a multi-temporal object classification and their accuracy were verified using reference data. The benefit of this thesis is finding generally applicable model useful to investigate the land cover change and evaluation of the potentially most appropriate seasonal configuration of images. Valuable is also methodology in this thesis which focus on selection of predictors and calculation the order of the most appropriate models, which is unique in the available literature. The thesis provides useful findings fitting to insufficiently explored issue of Change Detection arable land to grassland. Powered by TCPDF (www.tcpdf.org)

This thesis is concerned with data analysis, especially with principal component analysis and its sparse modi cation (SPCA), which is NP-hard-to- solve. SPCA problem can be recast into the regression framework in which spar- sity is usually induced with ℓ1-penalty. In the thesis, we propose to use iteratively reweighted ℓ2-penalty instead of the aforementioned ℓ1-approach. We compare the resulting algorithm with several well-known approaches to SPCA using both simulation study and interesting practical example in which we analyze voting re- cords of the Parliament of the Czech Republic. We show experimentally that the proposed algorithm outperforms the other considered algorithms. We also prove convergence of both the proposed algorithm and the original regression-based approach to PCA. vi

This cross-sectional study aimed to generate an exploratory functional model of deliberate self-harm (DSH) occurring among a South African student population (n = 603) using Principle Component Analysis (PCA). A battery of instruments, including the Deliberate Self-Harm Inventory (DSHI) and the Functional Assessment of Self-Mutilation (FASM) were administered to students. Following Nock and Prinstein’s (2004) four-factor model on the functions of DSH, self-reported reasons for DSH on the FASM were conceptualised in light of the reinforcement mechanisms of social learning theory. PCA revealed a four-factor model for the reasons students engaged in DSH, constituent of the following components: social positive reinforcement, social negative reinforcement, automatic reinforcement, and reinforcement regarding self-image. A unique factor manifests in the self-image reinforcement component, suggesting divergence in the functions of DSH between student and adolescent populations. / Dissertation (MA)--University of Pretoria, 2013. / gm2014 / Psychology / unrestricted

Magister Scientiae - MSc / Many hydrocarbon reservoirs are situated in barren sequences that display poor stratigraphic control. Correlation between the wells can become extremely difficult and traditional correlation techniques can prove to be inadequate. Past studies have shown that trace and major element concentrations can be used as a correlation tool. This practice of using geochemical fingerprints to characterize between wells is called Chemostratigraphic analysis. (Pearce et al, 1999) Chemostratigraphy has been recognized as a very important correlation technique as it can be used for rocks of any age, in any geological setting as well as sequences that are traditionally defined as barren. Chemostratigraphic analyses can be used as a means of getting rid of ambiguities within data produced by traditional correlation methods such as Biostratigraphy, Lithostratigraphy and Geophysical Logging. In areas where stratigraphic data is not available it can be used to construct correlation frameworks for the sequences found in the area. The motivation behind this study is that the research is not only worthy of academic investigation, but can also provide the industry with new insights into areas that were previously misunderstood because traditional correlation methods were not adequate. The study area, the Orange basin, is located offshore South Africa and is largely underexplored. The basin, that hosts two gas field namely the Ibhubesi and the Kudu gas fields, has large potential but in the past has not been given due attention with only 34 wells being drilled in the area. The Orange basin has recently been the topic of investigation because of the belief that it may be hosts to more hydrocarbons. This study will utilise Chemostratigraphy to attempt to provide geological information on this relatively under-explored basin. The aim of this research study is to produce a chemostratigraphic framework -scheme for the Orange Basin in order to facilitate reservoir scale interwell correlation. The Objectives of this research study will be to identify chemostratigraphic units or indices, to prove the adequate use of chemostratigraphy as an independent correlation technique and to integrate the chemostratigraphy and petrophysical characteristics of the four wells to facilitate lithological identification. / South Africa

This thesis aims to identify main driving market risk factors of different strategies implemented by hedge funds by looking at correlation coefficients, implementing Principal Component Analysis and analyzing "loadings" for first three principal components, which explain the largest portion of the variation of hedge funds' returns. In the next step, a stepwise regression through iteration process includes and excludes market risk factors for each strategy, searching for the combination of risk factors which will offer a model with the best "fit", based on The Akaike Information Criterion - AIC and Bayesian Information Criterion - BIC. Lastly, to avoid counterfeit results and overcome model uncertainty issues a Bayesian Model Average - BMA approach was taken. Key words: Hedge Funds, hedge funds' strategies, market risk, principal component analysis, stepwise regression, Akaike Information Criterion, Bayesian Information Criterion, Bayesian Model Averaging Author's e-mail: aleksaradosavcevic@gmail.com Supervisor's e-mail: mp.princ@seznam.cz

This dissertation introduces a new approach for assessing the effects of pediatric epilepsy on the language connectome. Two novel data-driven network construction approaches are presented. These methods rely on connecting different brain regions using either extent or intensity of language related activations as identified by independent component analysis of fMRI data. An auditory description decision task (ADDT) paradigm was used to activate the language network for 29 patients and 30 controls recruited from three major pediatric hospitals. Empirical evaluations illustrated that pediatric epilepsy can cause, or is associated with, a network efficiency reduction. Patients showed a propensity to inefficiently employ the whole brain network to perform the ADDT language task; on the contrary, controls seemed to efficiently use smaller segregated network components to achieve the same task. To explain the causes of the decreased efficiency, graph theoretical analysis was carried out. The analysis revealed no substantial global network feature differences between the patient and control groups. It also showed that for both subject groups the language network exhibited small-world characteristics; however, the patient’s extent of activation network showed a tendency towards more random networks. It was also shown that the intensity of activation network displayed ipsilateral hub reorganization on the local level. The left hemispheric hubs displayed greater centrality values for patients, whereas the right hemispheric hubs displayed greater centrality values for controls. This hub hemispheric disparity was not correlated with a right atypical language laterality found in six patients. Finally it was shown that a multi-level unsupervised clustering scheme based on self-organizing maps, a type of artificial neural network, and k-means was able to fairly and blindly separate the subjects into their respective patient or control groups. The clustering was initiated using the local nodal centrality measurements only. Compared to the extent of activation network, the intensity of activation network clustering demonstrated better precision. This outcome supports the assertion that the local centrality differences presented by the intensity of activation network can be associated with focal epilepsy.