Otimização de placas compósitas laminadas utilizando redes neuronais e enxames de partículas

Trabalho Final de mestrado para obtenção do grau de Mestre em engenharia Mecância

Parallelization of web processing services on cloud computing: A case study of Geostatistical Methods

Dissertation submitted in partial fulfillment of the requirements for the Degree of Master of Science in Geospatial Technologies.

Finite integration methods for isospectral flows

In this paper we consider the approximate computation of isospectral flows based on finite integration methods( FIM) with radial basis functions( RBF) interpolation,a new algorithm is developed. Our method ensures the symmetry of the solutions. Numerical experiments demonstrate that the solutions have higher accuracy by our algorithm than by the second order Runge- Kutta( RK2) method.

Automatic Mallampati Classification Using Active Appearance Models

Difficult tracheal intubation assessment is an important research topic in anesthesia as failed intubations are important causes of mortality in anesthetic practice. The modified Mallampati score is widely used, alone or in conjunction with other criteria, to predict the difficulty of intubation. This work presents an automatic method to assess the modified Mallampati score from an image of a patient with the mouth wide open. For this purpose we propose an active appearance models (AAM) based method and use linear support vector machines (SVM) to select a subset of relevant features obtained using the AAM. This feature selection step proves to be essential as it improves drastically the performance of classification, which is obtained using SVM with RBF kernel and majority voting. We test our method on images of 100 patients undergoing elective surgery and achieve 97.9% accuracy in the leave-one-out crossvalidation test and provide a key element to an automatic difficult intubation assessment system.

Contrast-enhanced ultrasound after devascularisation of neuroendocrine liver metastases: functional and morphological evaluation.

OBJECTIVE: To evaluate morphological and perfusion changes in liver metastases of neuroendocrine tumours by contrast-enhanced ultrasound (CEUS) after transarterial embolisation with bead block (TAE) or trans-arterial chemoembolisation with doxorubicin-eluting beads (DEB-TACE). METHODS: In this retrospective study, seven patients underwent TAE, and ten underwent DEB-TACE using beads of the same size. At 1 day before embolisation, 2 days, 1 month and 3 months after the procedure, a destruction-replenishment study using CEUS was performed with a microbubble-enhancing contrast material on a reference tumour. Relative blood flow (rBF) and relative blood volume (rBV) were obtained from the ratio of values obtained in the tumour and in adjacent liver parenchyma. Morphological parameters such as the tumour's major diameter and the viable tumour's major diameter were also measured. A parameter combining functional and morphological data, the tumour vitality index (TVI), was studied. The Wilcoxon rank-sum test and Fisher's test were used to compare treatment groups. RESULTS: At 3 months rBF, rBV and TVI were significantly lower (P = 0.005, P = 0.04 and P = 0.03) for the group with doxorubicin. No difference in morphological parameters was found throughout the follow-up. CONCLUSIONS: One parameter, TVI, could evaluate the morphological and functional response to treatments.

Thorax, pelvis and hip pattern in the frontal plane during walking in unilateral transtibial amputees: biomechanical analysis.

BACKGROUND Lower limb amputees exhibit postural control deficits during standing which can affect their walking ability. OBJECTIVES The primary purpose of the present study was to analyze the thorax, pelvis, and hip kinematics and the hip internal moment in the frontal plane during gait in subjects with Unilateral Transtibial Amputation (UTA). METHOD The participants included 25 people with UTA and 25 non-amputees as control subjects. Gait analysis was performed using the Vicon(r) Motion System. We analyzed the motion of the thorax, pelvis, and hip (kinematics) as well as the hip internal moment in the frontal plane. RESULTS The second peak of the hip abductor moment was significantly lower on the prosthetic side than on the sound side (p=.01) and the control side (right: p=.01; left: p=.01). During middle stance, the opposite side of the pelvis was higher on the prosthetic side compared to the control side (right: p=.01: left: p=.01). CONCLUSIONS The joint internal moment at the hip in the frontal plane was lower on the prosthetic side than on the sound side or the control side. Thorax and pelvis kinematics were altered during the stance phase on the prosthetic side, presumably because there are mechanisms which affect postural control during walking.

Renal effects of adenosine A1-receptor blockade with 8-cyclopentyl-1,3-dipropylxanthine in hypoxemic newborn rabbits.

The key role of intrarenal adenosine in mediating the hypoxemic acute renal insufficiency in newborn rabbits has been well demonstrated using the nonspecific adenosine antagonist theophylline. The present study was designed to define the role of adenosine A1 receptors during systemic hypoxemia by using the specific A1-receptor antagonist 8-cyclopentyl-1,3-dipropylxanthine (DPCPX). Renal function parameters were assessed in 31 anesthetized and mechanically ventilated newborn rabbits. In normoxia, DPCPX infusion induced a significant increase in diuresis (+44%) and GFR (+19%), despite a significant decrease in renal blood flow (RBF) (-22%) and an increase in renal vascular resistance (RVR) (+37%). In hypoxemic conditions, diuresis (-19%), GFR (-26%), and RBF (-35%) were decreased, whereas RVR increased (+33%). DPCPX administration hindered the hypoxemia-induced decrease in GFR and diuresis. However, RBF was still significantly decreased (-27%), whereas RVR increased (+22%). In all groups, the filtration fraction increased significantly. The overall results support the hypothesis that, in physiologic conditions, intrarenal adenosine plays a key role in regulating glomerular filtration in the neonatal period through preferential A1-mediated afferent vasoconstriction. During a hypoxemic stress, the A1-specific antagonist DPCPX only partially prevented the hypoxemia-induced changes, as illustrated by the elevated RVR and drop in RBF. These findings imply that the contribution of intrarenal adenosine to the acute adverse effects of hypoxemia might not be solely mediated via the A1 receptor.

PREVISÃO DA ENERGIA EÓLICA - SANTIAGO - CABO VERDE

Neste trabalho procura-se utilizar modelos de previsão de séries temporais para prever a produção da energia elétrica a partir da energia eólica em Cabo Verde, particularmente na ilha de Santiago. É um problema que tem recebido especial atenção dos pesquisadores nos últimos anos. Prever o futuro, e em especial o comportamento de séries temporais, é fundamental em análises e apoio à tomada de decisões, e continua sendo um desafio para a estatística e para computação. Foram utilizados modelos, Holt-Winters, ARIMA e redes neuronais artificiais, Função de Base Radial (RNAs-RBF) e Perceptron de múltiplas camadas (RNAs- MLP). O modelo Holt-Winters é um modelo de previsão exponencial, conhecido por lidar com elementos de tendência e sazonalidade de uma série temporal. O modelo ARIMA que possui apenas uma variável, descreve o comportamento de uma variável em termos de seus valores passados. As redes neurais têm-se mostrado grandes ferramentas na aplicação de previsões de séries temporais. Neste contexto, neste trabalho propõe-se a realização de uma análise comparativa desses modelos não-lineares para a previsão, tentando encontrar qual o modelo que melhor se adapta à série temporal. Todo o trabalho foi realizado com recurso ao programa estatístico R versão 3.0.1 (2013-05- 16)

Torasemide is an effective diuretic in the newborn rabbit.

The effect of intravenous (i.v.) torasemide on diuresis and renal function was evaluated in three groups of normoxemic, 5- to 10-day-old, newborn New Zealand White rabbits. The animals of group 1 received 0.2 mg/kg of torasemide i.v., whereas in group 2 an i.v. dose of 1.0 mg/kg was given. The third group of animals received a bolus i.v. dose of 1.0 mg/kg torasemide with continuous i.v. replacement of estimated urinary fluid and electrolyte losses. Torasemide proved to be an effective, potassium-sparing diuretic, without significant effect on glomerular filtration rate (GFR). Renal blood flow (RBF) fell and the renal vascular resistance (RVR) rose in all three groups of animals, although the rise in RVR in group 3 was not significant. These changes in renal hemodynamics were most pronounced in the animals of group 2 and are probably secondary to torasemide-induced hypovolemia (2.8% loss of body weight) and accompanying humoral reactions, such as an increase in angiotensin II (not measured). When the latter is prevented by simultaneous re-infusion of an electrolyte solution (group 3), replacing urinary losses, GFR increases and the changes in RBF and RVR are blunted. We conclude that torasemide is an effective, potassium-sparing diuretic in newborn rabbits. No evidence was found for a vasodilatory action of the drug.

Mult-class differentiation of cannabis seedlings in a forensic context

This article presents an experimental study about the classification ability of several classifiers for multi-classclassification of cannabis seedlings. As the cultivation of drug type cannabis is forbidden in Switzerland lawenforcement authorities regularly ask forensic laboratories to determinate the chemotype of a seized cannabisplant and then to conclude if the plantation is legal or not. This classification is mainly performed when theplant is mature as required by the EU official protocol and then the classification of cannabis seedlings is a timeconsuming and costly procedure. A previous study made by the authors has investigated this problematic [1]and showed that it is possible to differentiate between drug type (illegal) and fibre type (legal) cannabis at anearly stage of growth using gas chromatography interfaced with mass spectrometry (GC-MS) based on therelative proportions of eight major leaf compounds. The aims of the present work are on one hand to continueformer work and to optimize the methodology for the discrimination of drug- and fibre type cannabisdeveloped in the previous study and on the other hand to investigate the possibility to predict illegal cannabisvarieties. Seven classifiers for differentiating between cannabis seedlings are evaluated in this paper, namelyLinear Discriminant Analysis (LDA), Partial Least Squares Discriminant Analysis (PLS-DA), Nearest NeighbourClassification (NNC), Learning Vector Quantization (LVQ), Radial Basis Function Support Vector Machines(RBF SVMs), Random Forest (RF) and Artificial Neural Networks (ANN). The performance of each method wasassessed using the same analytical dataset that consists of 861 samples split into drug- and fibre type cannabiswith drug type cannabis being made up of 12 varieties (i.e. 12 classes). The results show that linear classifiersare not able to manage the distribution of classes in which some overlap areas exist for both classificationproblems. Unlike linear classifiers, NNC and RBF SVMs best differentiate cannabis samples both for 2-class and12-class classifications with average classification results up to 99% and 98%, respectively. Furthermore, RBFSVMs correctly classified into drug type cannabis the independent validation set, which consists of cannabisplants coming from police seizures. In forensic case work this study shows that the discrimination betweencannabis samples at an early stage of growth is possible with fairly high classification performance fordiscriminating between cannabis chemotypes or between drug type cannabis varieties.

Beneficial effect of insulin-like growth factor-1 on hypoxemic renal dysfunction in the newborn rabbit.

Acute normocapnic hypoxemia can cause functional renal insufficiency by increasing renal vascular resistance (RVR), leading to renal hypoperfusion and decreased glomerular filtration rate (GFR). Insulin-like growth factor 1 (IGF-1) activity is low in fetuses and newborns and further decreases during hypoxia. IGF-1 administration to humans and adult animals induces pre- and postglomerular vasodilation, thereby increasing GFR and renal blood flow (RBF). A potential protective effect of IGF-1 on renal function was evaluated in newborn rabbits with hypoxemia-induced renal insufficiency. Renal function and hemodynamic parameters were assessed in 17 anesthetized and mechanically ventilated newborn rabbits. After hypoxemia stabilization, saline solution (time control) or IGF-1 (1 mg/kg) was given as an intravenous (i.v.) bolus, and renal function was determined for six 30-min periods. Normocapnic hypoxemia significantly increased RVR (+16%), leading to decreased GFR (-14%), RBF (-19%) and diuresis (-12%), with an increased filtration fraction (FF). Saline solution resulted in a worsening of parameters affected by hypoxemia. Contrarily, although mean blood pressure decreased slightly but significantly, IGF-1 prevented a further increase in RVR, with subsequent improvement of GFR, RBF and diuresis. FF indicated relative postglomerular vasodilation. Although hypoxemia-induced acute renal failure was not completely prevented, IGF-1 elicited efferent vasodilation, thereby precluding a further decline in renal function.

Complementary effects of adenosine and angiotensin II in hypoxemia-induced renal dysfunction in the rabbit.

The acute renal effects of hypoxemia and the ability of the co-administration of an angiotensin converting enzyme inhibitor (perindoprilat) and an adenosine receptor antagonist (theophylline) to prevent these effects were assessed in anesthetized and mechanically-ventilated rabbits. Renal blood flow (RBF) and glomerular filtration rate (GFR) were determined by the clearances of para-aminohippuric acid and inulin, respectively. Each animal acted as its own control. In 8 untreated rabbits, hypoxemia induced a significant drop in mean blood pressure (-12 +/- 2%), GFR (-16 +/- 3%) and RBF (-12 +/- 3%) with a concomitant increase in renal vascular resistance (RVR) (+ 18 +/- 5%), without changes in filtration fraction (FF) (-4 +/- 2%). These results suggest the occurrence of both pre- and postglomerular vasoconstriction during the hypoxemic stress. In 7 rabbits pretreated with intravenous perindoprilat (20 microg/kg), the hypoxemia-induced changes in RBF and RVR were prevented. FF decreased significantly (-18 +/- 2%), while the drop in GFR was partially blunted. These results could be explained by the inhibition of the angiotensin-mediated efferent vasoconstriction by perindoprilat. In 7 additional rabbits, co-administration of perindoprilat and theophylline (1 mg/kg) completely prevented the hypoxemia-induced changes in RBF (+ 11 +/- 3%) and GFR (+ 2 +/- 3%), while RVR decreased significantly (-14 +/- 3%). Since adenosine and angiotensin II were both shown to participate, at least in part, in the renal changes induced by hypoxemia, the beneficial effects of perindoprilat and theophylline in this model could be mediated by complementary actions of angiotensin II and adenosine on the renal vasculature.

Machine learning for geospatial data : algorithms, software tools and case studies

Résumé Cette thèse est consacrée à l'analyse, la modélisation et la visualisation de données environnementales à référence spatiale à l'aide d'algorithmes d'apprentissage automatique (Machine Learning). L'apprentissage automatique peut être considéré au sens large comme une sous-catégorie de l'intelligence artificielle qui concerne particulièrement le développement de techniques et d'algorithmes permettant à une machine d'apprendre à partir de données. Dans cette thèse, les algorithmes d'apprentissage automatique sont adaptés pour être appliqués à des données environnementales et à la prédiction spatiale. Pourquoi l'apprentissage automatique ? Parce que la majorité des algorithmes d'apprentissage automatiques sont universels, adaptatifs, non-linéaires, robustes et efficaces pour la modélisation. Ils peuvent résoudre des problèmes de classification, de régression et de modélisation de densité de probabilités dans des espaces à haute dimension, composés de variables informatives spatialisées (« géo-features ») en plus des coordonnées géographiques. De plus, ils sont idéaux pour être implémentés en tant qu'outils d'aide à la décision pour des questions environnementales allant de la reconnaissance de pattern à la modélisation et la prédiction en passant par la cartographie automatique. Leur efficacité est comparable au modèles géostatistiques dans l'espace des coordonnées géographiques, mais ils sont indispensables pour des données à hautes dimensions incluant des géo-features. Les algorithmes d'apprentissage automatique les plus importants et les plus populaires sont présentés théoriquement et implémentés sous forme de logiciels pour les sciences environnementales. Les principaux algorithmes décrits sont le Perceptron multicouches (MultiLayer Perceptron, MLP) - l'algorithme le plus connu dans l'intelligence artificielle, le réseau de neurones de régression généralisée (General Regression Neural Networks, GRNN), le réseau de neurones probabiliste (Probabilistic Neural Networks, PNN), les cartes auto-organisées (SelfOrganized Maps, SOM), les modèles à mixture Gaussiennes (Gaussian Mixture Models, GMM), les réseaux à fonctions de base radiales (Radial Basis Functions Networks, RBF) et les réseaux à mixture de densité (Mixture Density Networks, MDN). Cette gamme d'algorithmes permet de couvrir des tâches variées telle que la classification, la régression ou l'estimation de densité de probabilité. L'analyse exploratoire des données (Exploratory Data Analysis, EDA) est le premier pas de toute analyse de données. Dans cette thèse les concepts d'analyse exploratoire de données spatiales (Exploratory Spatial Data Analysis, ESDA) sont traités selon l'approche traditionnelle de la géostatistique avec la variographie expérimentale et selon les principes de l'apprentissage automatique. La variographie expérimentale, qui étudie les relations entre pairs de points, est un outil de base pour l'analyse géostatistique de corrélations spatiales anisotropiques qui permet de détecter la présence de patterns spatiaux descriptible par une statistique. L'approche de l'apprentissage automatique pour l'ESDA est présentée à travers l'application de la méthode des k plus proches voisins qui est très simple et possède d'excellentes qualités d'interprétation et de visualisation. Une part importante de la thèse traite de sujets d'actualité comme la cartographie automatique de données spatiales. Le réseau de neurones de régression généralisée est proposé pour résoudre cette tâche efficacement. Les performances du GRNN sont démontrées par des données de Comparaison d'Interpolation Spatiale (SIC) de 2004 pour lesquelles le GRNN bat significativement toutes les autres méthodes, particulièrement lors de situations d'urgence. La thèse est composée de quatre chapitres : théorie, applications, outils logiciels et des exemples guidés. Une partie importante du travail consiste en une collection de logiciels : Machine Learning Office. Cette collection de logiciels a été développée durant les 15 dernières années et a été utilisée pour l'enseignement de nombreux cours, dont des workshops internationaux en Chine, France, Italie, Irlande et Suisse ainsi que dans des projets de recherche fondamentaux et appliqués. Les cas d'études considérés couvrent un vaste spectre de problèmes géoenvironnementaux réels à basse et haute dimensionnalité, tels que la pollution de l'air, du sol et de l'eau par des produits radioactifs et des métaux lourds, la classification de types de sols et d'unités hydrogéologiques, la cartographie des incertitudes pour l'aide à la décision et l'estimation de risques naturels (glissements de terrain, avalanches). Des outils complémentaires pour l'analyse exploratoire des données et la visualisation ont également été développés en prenant soin de créer une interface conviviale et facile à l'utilisation. Machine Learning for geospatial data: algorithms, software tools and case studies Abstract The thesis is devoted to the analysis, modeling and visualisation of spatial environmental data using machine learning algorithms. In a broad sense machine learning can be considered as a subfield of artificial intelligence. It mainly concerns with the development of techniques and algorithms that allow computers to learn from data. In this thesis machine learning algorithms are adapted to learn from spatial environmental data and to make spatial predictions. Why machine learning? In few words most of machine learning algorithms are universal, adaptive, nonlinear, robust and efficient modeling tools. They can find solutions for the classification, regression, and probability density modeling problems in high-dimensional geo-feature spaces, composed of geographical space and additional relevant spatially referenced features. They are well-suited to be implemented as predictive engines in decision support systems, for the purposes of environmental data mining including pattern recognition, modeling and predictions as well as automatic data mapping. They have competitive efficiency to the geostatistical models in low dimensional geographical spaces but are indispensable in high-dimensional geo-feature spaces. The most important and popular machine learning algorithms and models interesting for geo- and environmental sciences are presented in details: from theoretical description of the concepts to the software implementation. The main algorithms and models considered are the following: multi-layer perceptron (a workhorse of machine learning), general regression neural networks, probabilistic neural networks, self-organising (Kohonen) maps, Gaussian mixture models, radial basis functions networks, mixture density networks. This set of models covers machine learning tasks such as classification, regression, and density estimation. Exploratory data analysis (EDA) is initial and very important part of data analysis. In this thesis the concepts of exploratory spatial data analysis (ESDA) is considered using both traditional geostatistical approach such as_experimental variography and machine learning. Experimental variography is a basic tool for geostatistical analysis of anisotropic spatial correlations which helps to understand the presence of spatial patterns, at least described by two-point statistics. A machine learning approach for ESDA is presented by applying the k-nearest neighbors (k-NN) method which is simple and has very good interpretation and visualization properties. Important part of the thesis deals with a hot topic of nowadays, namely, an automatic mapping of geospatial data. General regression neural networks (GRNN) is proposed as efficient model to solve this task. Performance of the GRNN model is demonstrated on Spatial Interpolation Comparison (SIC) 2004 data where GRNN model significantly outperformed all other approaches, especially in case of emergency conditions. The thesis consists of four chapters and has the following structure: theory, applications, software tools, and how-to-do-it examples. An important part of the work is a collection of software tools - Machine Learning Office. Machine Learning Office tools were developed during last 15 years and was used both for many teaching courses, including international workshops in China, France, Italy, Ireland, Switzerland and for realizing fundamental and applied research projects. Case studies considered cover wide spectrum of the real-life low and high-dimensional geo- and environmental problems, such as air, soil and water pollution by radionuclides and heavy metals, soil types and hydro-geological units classification, decision-oriented mapping with uncertainties, natural hazards (landslides, avalanches) assessments and susceptibility mapping. Complementary tools useful for the exploratory data analysis and visualisation were developed as well. The software is user friendly and easy to use.

Diversity and genetic struture in cajá tree (spondias mombin l.) populations in northeastern Brazil

Spondias mombin L. is a fruit tree from the American continent from the Anacardiaceae family. In Brazil it is common in different vegetation types but is more frequent in the Atlantic and Amazonian rainforests. It is economically important because of its fruits, which are widely consumed raw or processed as fruit jellies, juices and ice creams. The leaves have great importance in the pharmaceutical industry because of their antibacterial properties. In the state of Pernambuco, cajá tree is widely distributed in the Zona da Mata region and less frequently in the Agreste and Sertão areas. In this work diversity and genetic structure were studied in four populations of cajá tree from Pernambuco's Zona da Mata, Northeast Brazil, using isozymes polymorphism analyses from electrophoreses. The result showed 100% of polymorphism (P) for nine alleles and the average of alleles per locus s was 2.4. The expected heterozygosity ranged from 0.530 to 0.574 and the observed heterozygosity , from 0.572 to 0.735. It was not observed inbreeding and the average F IT was -0.175, whereas within population inbreeding (f) varied from -0.08 to- 0.37. The genetic divergence among the populations (F ST) ranged from 0.006 to 0.028 and the average was 0.026. The average of estimated gene flow (Nm) was high (5.27). The CG-IPA population, corresponding to the germplasm collection of IPA, showed more than 96% of genetic similarity with other populations; therefore, it is a good representative of the existent genetic diversity in the Zona da Mata region.

Diversity and genetic structure in natural populations of Hancornia speciosa var. speciosa Gomes in northeastern Brazil

Hancornia speciosa Gomes is a fruit tree native from Brazil that belongs to Apocinaceae family, and is popularly known as Mangabeira. Its fruits are widely consumed raw or processed as fruit jam, juices and ice creams, which have made it a target of intense exploitation. The extractive activities and intense human activity on the environment of natural occurrence of H. speciosa has caused genetic erosion in the species and little is known about the ecology or genetic structure of natural populations. The objective of this research was the evaluation of the genetic diversity and genetic structure of H. speciosa var. speciosa. The genetic variability was assessed using 11 allozyme loci with a sample of 164 individuals distributed in six natural populations located in the States of Pernambuco and Alagoas, Northeastern Brazil. The results showed a high level of genetic diversity within the species (e= 0.36) seeing that the most of the genetic variability of H. speciosa var. speciosa is within its natural populations with low difference among populations ( or = 0.081). The inbreeding values within ( = -0.555) and among populations ( =-0.428) were low showing lacking of endogamy and a surplus of heterozygotes. The estimated gene flow ( m ) was high, ranging from 2.20 to 13.18, indicating to be enough to prevent the effects of genetic drift and genetic differentiation among populations. The multivariate analyses indicated that there is a relationship between genetic and geographical distances, which was confirmed by a spatial pattern analysis using Mantel test (r = 0.3598; p = 0.0920) with 1000 random permutations. The high genetic diversity index in these populations indicates potential for in situ genetic conservation.