L’Imagerie par résonance magnétique (IRM) est un outil fondamental pour l’exploration in vivo du développement du cerveau chez le fœtus, le bébé et l’enfant. Elle fournit plusieurs paramètres quantitatifs qui reflètent les changements des propriétés tissulaires au cours du développement en fonction de différents processus de maturation. Cependant, l’évaluation fiable de la maturation de la substance blanche est encore une question ouverte: d'une part, aucun de ces paramètres ne peut décrire toute la complexité des changements sous-jacents; d'autre part, aucun d'eux n’est spécifique d’un processus de développement ou d’une propriété tissulaire particulière. L’implémentation d’approches multiparamétriques combinant les informations complémentaires issues des différents paramètres IRM devrait permettre d’améliorer notre compréhension du développement du cerveau. Dans ce travail de thèse, je présente deux exemples de telles approches et montre leur pertinence pour l'étude de la maturation des faisceaux de substance blanche. La première approche fournit une mesure globale de la maturation basée sur la distance de Mahalanobis calculée à partir des différents paramètres IRM (temps de relaxation T1 et T2, diffusivités longitudinale et transverse du tenseur de diffusion DTI) chez des nourrissons (âgés de 3 à 21 semaines) et des adultes. Cette approche offre une meilleure description de l’asynchronisme de maturation à travers les différents faisceaux que les approches uniparamétriques. De plus, elle permet d'estimer les délais relatifs de maturation entre faisceaux. La seconde approche vise à quantifier la myélinisation des tissus cérébraux, en calculant la fraction de molécules d’eau liées à la myéline (MWF) en chaque voxel des images. Cette approche est basée sur un modèle tissulaire avec trois composantes ayant des caractéristiques de relaxation spécifiques, lesquelles ont été pré-calibrées sur trois jeunes adultes sains. Elle permet le calcul rapide des cartes MWF chez les nourrissons et semble bien révéler la progression de la myélinisation à l’échelle cérébrale. La robustesse de cette approche a également été étudiée en simulations. Une autre question cruciale pour l'étude du développement de la substance blanche est l'identification des faisceaux dans le cerveau des enfants. Dans ce travail de thèse, je décris également la création d'un atlas préliminaire de connectivité structurelle chez des enfants âgés de 17 à 81 mois, permettant l'extraction automatique des faisceaux à partir des données de tractographie. Cette approche a démontré sa pertinence pour l'évaluation régionale de la maturation de la substance blanche normale chez l’enfant. Pour finir, j’envisage dans la dernière partie du manuscrit les applications potentielles des différentes méthodes précédemment décrites pour l’étude fine des réseaux de substance blanche dans le cadre de deux exemples spécifiques de pathologies : les épilepsies focales et la leucodystrophie métachromatique. / Magnetic Resonance Imaging (MRI) is a fundamental tool for in vivo investigation of brain development in newborns, infants and children. It provides several quantitative parameters that reflect changes in tissue properties during development depending on different undergoing maturational processes. However, reliable evaluation of the white matter maturation is still an open question: on one side, none of these parameters can describe the whole complexity of the undergoing changes; on the other side, neither of them is specific to any particular developmental process or tissue property. Developing multiparametric approaches combining complementary information from different MRI parameters is expected to improve our understanding of brain development. In this PhD work, I present two examples of such approaches and demonstrate their relevancy for investigation of maturation across different white matter bundles. The first approach provides a global measure of maturation based on the Mahalanobis distance calculated from different MRI parameters (relaxation times T1 and T2, longitudinal and transverse diffusivities from Diffusion Tensor Imaging, DTI) in infants (3-21 weeks) and adults. This approach provides a better description of the asynchronous maturation across the bundles than univariate approaches. Furthermore, it allows estimating the relative maturational delays between the bundles. The second approach aims at quantifying myelination of brain tissues by calculating Myelin Water Fraction (MWF) in each image voxel. This approach is based on a 3-component tissue model, with each model component having specific relaxation characteristics that were pre-calibrated in three healthy adult subjects. This approach allows fast computing of the MWF maps from infant data and could reveal progression of the brain myelination. The robustness of this approach was further investigated using computer simulations. Another important issue for studying white matter development in children is bundles identification. In the last part of this work I also describe creation of a preliminary atlas of white matter structural connectivity in children aged 17-81 months. This atlas allows automatic extraction of the bundles from tractography datasets. This approach demonstrated its relevance for evaluation of regional maturation of normal white matter in children. Finally, in the last part of the manuscript I describe potential future applications of the previously developed methods to investigation of the white matter in cases of two specific pathologies: focal epilepsy and metachromatic leukodystrophy.

Conselho Nacional de Desenvolvimento Científico e Tecnológico / Many fatal traffic accidents are caused by fatigued and drowsy drivers. In this context, automatic drowsiness detection devices are an alternative to minimize this issue. In this work, two new methodologies to drowsiness detection are presented, considering a signal obtained from a single electroencephalography channel: (i) drowsiness detection through best m-term approximation, applied to the wavelet expansion of the analysed signal; (ii) drowsiness detection through Mahalanobis distance with wavelet coefficients. The results of both methodologies are compared with a method which uses Mahalanobis distance and Fourier coefficients to drowsiness detection. All methodologies consider the medical evaluation of the brain signal, given by the hypnogram, as a reference. / A sonolência diurna em motoristas, principal consequência da privação de sono, tem sido a causa de diversos acidentes graves de trânsito. Neste contexto, a utilização de dispositivos que alertem o condutor ao detectar automaticamente o estado de sonolência é uma alternativa para a minimização deste problema. Neste trabalho, duas novas metodologias para a detecção automática da sonolência são apresentadas, utilizando um único canal de eletroencefalografia para a obtenção do sinal: (i) detecção da sonolência via melhor aproximação por m-termos, aplicada aos coeficientes wavelets da expansão em série do sinal; e (ii) detecção da sonolência via distância de Mahalanobis e coeficientes wavelets. Os resultados de ambas as metodologias são comparados a uma implementação utilizando distância de Mahalanobis e coeficientes de Fourier. Para todas as metodologias, utiliza-se como referência a avaliação médica do sinal cerebral, dada pelo hipnograma.

Cluster and discriminant analysis belong to basic classification methods. Using cluster analysis can be a disordered group of objects organized into several internally homogeneous classes or clusters. Discriminant analysis creates knowledge based on the jurisdiction of existing classes classification rule, which can be then used for classifying units with an unknown group membership. The aim of this thesis is a comparison of discriminant analysis and different methods of cluster analysis. To reflect the distances between objects within each cluster, squeared Euclidean and Mahalanobis distances are used. In total, there are 28 datasets analyzed in this thesis. In case of leaving correlated variables in the set and applying squared Euclidean distance, Ward´s method classified objects into clusters the most successfully (42,0 %). After changing metrics on the Mahalanobis distance, the most successful method has become the furthest neighbor method (37,5 %). After removing highly correlated variables and applying methods with Euclidean metric, Ward´s method was again the most successful in classification of objects (42,0%). From the result implies that cluster analysis is more precise when excluding correlated variables than when leaving them in a dataset. The average result of discriminant analysis for data with correlated variables and also without correlated variables is 88,7 %.

This master's thesis deals with the problems of anomalies detection in data from clinical trials and medical registries. The purpose of this work is to perform literary research about quality of data in clinical trials and to design a personal algorithm for detection of anomalous records based on machine learning methods in real clinical data from current or completed clinical trials or medical registries. In the practical part is described the implemented algorithm of detection, consists of several parts: import of data from information system, preprocessing and transformation of imported data records with variables of different data types into numerical vectors, using well known statistical methods for detection outliers and evaluation of the quality and accuracy of the algorithm. The result of creating the algorithm is vector of parameters containing anomalies, which has to make the work of data manager easier. This algorithm is designed for extension the palette of information system functions (CLADE-IS) on automatic monitoring the quality of data by detecting anomalous records.

This master's thesis deals with the problems of anomalies detection in data from clinical trials and medical registries. The purpose of this work is to perform literary research about quality of data in clinical trials and to design a personal algorithm for detection of anomalous records based on machine learning methods in real clinical data from current or completed clinical trials or medical registries. In the practical part is described the implemented algorithm of detection, consists of several parts: import of data from information system, preprocessing and transformation of imported data records with variables of different data types into numerical vectors, using well known statistical methods for detection outliers and evaluation of the quality and accuracy of the algorithm. The result of creating the algorithm is vector of parameters containing anomalies, which has to make the work of data manager easier. This algorithm is designed for extension the palette of information system functions (CLADE-IS) on automatic monitoring the quality of data by detecting anomalous records.

Les analyses effectuées dans le cadre de ce mémoire ont été réalisées à l'aide du module MatchIt disponible sous l’environnent d'analyse statistique R. / Statistical analyzes of this thesis were performed using the MatchIt package available in the statistical analysis environment R. / L’estimation sans biais de l’effet causal d’une intervention nécessite la comparaison de deux groupes homogènes. Il est rare qu’une étude observationnelle dispose de groupes comparables et même une étude expérimentale peut se retrouver avec des groupes non comparables. Les chercheurs ont alors recours à des techniques de correction afin de rendre les deux groupes aussi semblables que possible. Le problème consiste alors à choisir la méthode de correction appropriée. En ce qui nous concerne, nous limiterons nos recherches à une famille de méthodes dites d’appariement. Il est reconnu que ce qui importe lors d’un appariement est l’équilibre des deux groupes sur les caractéristiques retenues. Autrement dit, il faut que les variables soient distribuées de façon similaire dans les deux groupes. Avant même de considérer la distribution des variables entre les deux groupes, il est nécessaire de savoir si les données en question permettent une inférence causale. Afin de présenter le problème de façon rigoureuse, le modèle causal contrefactuel sera exposé. Par la suite, les propriétés formelles de trois méthodes d’appariement seront présentées. Ces méthodes sont l’appariement par la distance de Mahalanobis, de l’appariement par le score de propension et de l’appariement génétique. Le choix de la technique d’appariement appropriée reposera sur quatre critères empiriques dont le plus important est la différence des moyennes standardisées. Les résultats obtenus à l’aide des données de l’Enquête longitudinale et expérimentale de Montréal (ÉLEM) indiquent que des trois techniques d’appariement, l’appariement génétique est celui qui équilibre mieux les variables entre les groupes sur tous les critères retenus. L’estimation de l’effet de l’intervention varie sensiblement d’une technique à l’autre, bien que dans tous les cas cet effet est non significatif. Ainsi, le choix d’une technique d’appariement influence l’estimation de l’effet d’une intervention. Il est donc impérieux de choisir la technique qui permet d’obtenir un équilibre optimal des variables selon les données à la disposition du chercheur. / The unbiased estimate of the causal effect of an intervention requires the comparison of two homogeneous groups. It is rare that an observational study has comparable groups and even an experiment may end up with non-comparable groups. The researchers then used correction techniques to make the two groups as similar as possible. The problem then is to choose the appropriate correction method. In our case, we will restrict our research to a family of so-called matching methods. It is recognized that what matters in a match is the balance between the two groups on selected characteristics. In other words, it is necessary that the variables are distributed similarly in both groups. Even before considering the distribution of variables between the two groups, it is necessary to know whether the data in question allow for causal inference. To present the problem rigorously, the counterfactual causal model will be exposed. Thereafter, the formal properties of three matching methods will be presented. Those methods are the Mahalanobis matching, the propensity score matching and genetic matching. The choice of the appropriate matching technique is based on four empirical criteria which the most important is the standardized mean difference. Results obtained using data from the Montréal Longitudinal and Experimental Study indicate that of the three matching techniques, genetic matching is the one that better balance the variables between groups on all criteria. The estimate of the effect of intervention varies substantially from one technique to another, although in all cases this effect is non significant. Thus, the selection of a matching technique influences the estimation of the effect of an intervention. Therefore, it is imperative to choose the technique that provides an optimal balance of the variables based on data available to the researcher.

This thesis investigates the problem of robot navigation using only landmark bearings. The proposed system allows a robot to move to a ground target location specified by the sensor values observed at this ground target posi- tion. The control actions are computed based on the difference between the current landmark bearings and the target landmark bearings. No Cartesian coordinates with respect to the ground are computed by the control system. The robot navigates using solely information from the bearing sensor space. Most existing robot navigation systems require a ground frame (2D Cartesian coordinate system) in order to navigate from a ground point A to a ground point B. The commonly used sensors such as laser range scanner, sonar, infrared, and vision do not directly provide the 2D ground coordi- nates of the robot. The existing systems use the sensor measurements to localise the robot with respect to a map, a set of 2D coordinates of the objects of interest. It is more natural to navigate between the points in the sensor space corresponding to A and B without requiring the Cartesian map and the localisation process. Research on animals has revealed how insects are able to exploit very limited computational and memory resources to successfully navigate to a desired destination without computing Cartesian positions. For example, a honeybee balances the left and right optical flows to navigate in a nar- row corridor. Unlike many other ants, Cataglyphis bicolor does not secrete pheromone trails in order to find its way home but instead uses the sun as a compass to keep track of its home direction vector. The home vector can be inaccurate, so the ant also uses landmark recognition. More precisely, it takes snapshots and compass headings of some landmarks. To return home, the ant tries to line up the landmarks exactly as they were before it started wandering. This thesis introduces a navigation method based on reflex actions in sensor space. The sensor vector is made of the bearings of some landmarks, and the reflex action is a gradient descent with respect to the distance in sensor space between the current sensor vector and the target sensor vec- tor. Our theoretical analysis shows that except for some fully characterized pathological cases, any point is reachable from any other point by reflex action in the bearing sensor space provided the environment contains three landmarks and is free of obstacles. The trajectories of a robot using reflex navigation, like other image- based visual control strategies, do not correspond necessarily to the shortest paths on the ground, because the sensor error is minimized, not the moving distance on the ground. However, we show that the use of a sequence of waypoints in sensor space can address this problem. In order to identify relevant waypoints, we train a Self Organising Map (SOM) from a set of observations uniformly distributed with respect to the ground. This SOM provides a sense of location to the robot, and allows a form of path planning in sensor space. The navigation proposed system is analysed theoretically, and evaluated both in simulation and with experiments on a real robot.

Android is the second most targeted operating system for malware authors and to counter the development of Android malware, more knowledge about their behavior is needed. There are mainly two approaches to analyze Android malware, namely static and dynamic analysis. Recently in 2017, a study and well labeled dataset, named AMD (Android Malware Dataset), consisting of over 24,000 malware samples was released. It is divided into 135 varieties based on similar malicious behavior, retrieved through static analysis of the file classes.dex in the APK of each malware, whereas the labeled features were determined by manual inspection of three samples in each variety. However, static analysis is known to be weak against obfuscation techniques, such as repackaging or dynamic loading, which can be exploited to avoid the analysis. In this study the second approach is utilized and all malware in the dataset are analyzed at run-time in order to monitor their dynamic behavior. However, analyzing malware at run-time has known weaknesses as well, as it can be avoided through, for instance, anti-emulator techniques. Therefore, the study aimed to explore the available sandbox environments for dynamic analysis, study the effectiveness of fingerprinting Android malware using one of the tools and investigate whether static features from AMD and the dynamic analysis correlate. For instance, by an attempt to classify the samples based on similar dynamic features and calculating the Pearson Correlation Coefficient (r) for all combinations of features from AMD and the dynamic analysis. The comparison of tools for dynamic analysis, showed a need of development, as most popular tools has been released for a long time and the common factor is a lack of continuous maintenance. As a result, the choice of sandbox environment for this study ended up as Droidbox, because of aspects like ease of use/install and easily adaptable for large scale analysis. Based on the dynamic features extracted with Droidbox, it could be shown that Android malware are more similar to the varieties which they belong to. The best metric for classifying samples to varieties, out of four investigated metrics, turned out to be Cosine Similarity, which received an accuracy of 83.6% for the entire dataset. The high accuracy indicated a correlation between the dynamic features and static features which the varieties are based on. Furthermore, the Pearson Correlation Coefficient confirmed that the manually extracted features, used to describe the varieties, and the dynamic features are correlated to some extent, which could be partially confirmed by a manual inspection in the end of the study.

Classically, wide-area synchronized power system monitoring has been an expensive task requiring significant investment in utility communications infrastructures for the service of relatively few costly sensors. The purpose of this research is to demonstrate the viability of power system monitoring from very low voltage levels (120 V). Challenging the accepted norms in power system monitoring, the document will present the use of inexpensive GPS time synchronized sensors in mass numbers at the distribution level. In the past, such low level monitoring has been overlooked due to a perceived imbalance between the required investment and the usefulness of the resulting deluge of information. However, distribution level monitoring offers several advantages over bulk transmission system monitoring. First, practically everyone with access to electricity also has a measurement port into the electric power system. Second, internet access and GPS availability have become pedestrian commodities providing a communications and synchronization infrastructure for the transmission of low-voltage measurements. Third, these ubiquitous measurement points exist in an interconnected fashion irrespective of utility boundaries. This work offers insight into which parameters are meaningful to monitor at the distribution level and provides applications that add unprecedented value to the data extracted from this level. System models comprising the entire Eastern Interconnection are exploited in conjunction with a bounty of distribution level measurement data for the development of wide-area disturbance detection, classification, analysis, and location routines. The main contributions of this work are fivefold: the introduction of a novel power system disturbance detection algorithm; the development of a power system oscillation damping analysis methodology; the development of several parametric and non-parametric power system disturbance location methods, new methods of power system phenomena visualization, and the proposal and mapping of an online power system event reporting scheme. / Ph. D.

在迴歸分析中，當資料中存在很多離群值時，偵測的工作變得非常不容易。 在此狀況下，我們無法使用傳統的殘差分析正確地偵測出其是否存在，此現象稱為遮蔽效應(The Masking Effect)。 而為了避免此效應的發生，我們利用最小中位數穩健迴歸估計值(Least Median Squares Estimator)正確地找出這些群集離群值，此估計值擁有最大即50﹪的容離值 (Breakdown point)。 在這篇論文中，用來求出最小中位數穩健迴歸估計值的演算法稱為步進搜尋演算法 (the Forward Search Algorithm)。 結果顯示，我們可以利用此演算法得到的穩健迴歸估計值，很快並有效率的找出資料中的群集離群值；另外，更進一步的結果顯示，我們只需從資料中隨機選取一百次子集，並進行步進搜尋，即可得到概似的穩健迴歸估計值並正確的找出那些群集離群值。 最後，我們利用鐘乳石圖(Stalactite Plot)列出所有被偵測到的離群值。 在多變量資料中，我們若使用Mahalanobis距離也會遭遇到同樣的屏蔽效應。 而此一問題，隨著另一高度穩健估計值的採用，亦可迎刃而解。 此估計值稱為最小體積橢圓體估計值 (Minimum Volume Ellipsoid)，其亦擁有最大即50﹪的容離值。 在此，我們也利用步進搜尋法求出此估計值，並利用鐘乳石圖列出所有被偵測到的離群值。 這篇論文的第二部分則利用變數轉換的技巧將迴歸資料中的殘差項常態化並且加強其等變異的特性以利後續的資料分析。 在步進搜尋進行的過程中，我們觀察分數統計量(Score Statistic)和其他相關診斷統計量的變化。 結果顯示，這些統計量一起提供了有關轉換參數選取豐富的資訊，並且我們亦可從步進搜尋進行的過程中觀察出某些離群值對參數選取的影響。 / Detecting regression outliers is not trivial when there are many of them. The methods of using classical diagnostic plots sometimes fail to detect them. This phenomenon is known as the masking effect. To avoid this, we propose to find out those multiple outliers by using a highly robust regression estimator called the least median squares (LMS) estimator which has maximal breakdown point. The algorithm in search of the LMS estimator is called the forward search algorithm. The estimator found by the forward search is shown to lead to the rapid detection of multiple outliers. Furthermore, the result reveals that 100 repeats of a simple forward search from a random starting subset are shown to provide sufficiently robust parameter estimators to reveal multiple outliers. Finally, those detected outliers are exhibited by the stalactite plot that shows greatly stable pattern of them. Referring to multivariate data, the Mahalanobis distance also suffers from the masking effect that can be remedied by using a highly robust estimator called the minimum volume ellipsoid (MVE) estimator. It can also be found by using the forward search algorithm and it also has maximal breakdown point. The detected outliers are then displayed in the stalactite plot. The second part of this dissertation is the transformation of regression data so that the approximate normality and the homogeneity of the residuals can be achieved. During the process of the forward search, we monitor the quantity of interest called score statistic and some other diagnostic plots. They jointly provide a wealth of information about transformation along with the effect of individual observation on this statistic.