Diabetic retinopathy (DR) is becoming a global health concern, which causes the loss of vision of most patients with the disease. Due to the vast prevalence of the disease, the automated detection of the DR is needed for quick diagnoses where the progress of the disease is monitored by detection of exudates changes and their classifications in the fundus retina images. Today in the automated system of the disease diagnoses, several image enhancement methods are used on original Fundus images. The primary goal of this thesis is to make a comparison of three of popular enhancement methods of the Mahalanobis Distance (MD), the Histogram Equalization (HE) and the Contrast Limited Adaptive Histogram Equalization (CLAHE). By quantifying the comparison in the aspect of the ability to detect and classify exudates, the best of the three enhancement methods is implemented to detect and classify soft and hard exudates. A graphical user interface is also adopted, with the help of MATLAB. The results showed that the MD enhancement method yielded better results in enhancement of the digital images compared to the HE and the CLAHE. The technique also enabled this study to successfully classify exudates into hard and soft exudates classification. Generally, the research concluded that the method that was suggested yielded the best results regarding the detection of the exudates; its classification and management can be suggested to the doctors and the ophthalmologists.

Weather radar imagery is important for several remote sensing applications including tracking of storm fronts and radar echo classification. In particular, tracking of precipitation events is useful for both forecasting and classification of rain/non-rain events since non-rain events usually appear to be static compared to rain events. Recent weather radar imaging-based forecasting approaches [3] consider that precipitation events can be modeled as a combination of localized functions using Radial Basis Function Neural Networks (RBFNNs). Tracking of rain events can be performed by tracking the parameters of these localized functions. The RBFNN-based techniques used in forecasting are not only computationally expensive, but also moderately effective in modeling small size precipitation events. In this thesis, an existing RBFNN technique [3] was implemented to verify its computational efficiency and forecasting effectiveness. The feasibility of modeling precipitation events using RBFNN effectively was evaluated, and several modifications to the existing technique have been proposed.

Made available in DSpace on 2017-07-10T17:36:53Z (GMT). No. of bitstreams: 1 2014_Diss_Jonas_Francisco_Egewarth.pdf: 723125 bytes, checksum: 682d564ae8329cf87ce6675a399a243d (MD5) Previous issue date: 2014-02-26 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The present study was conducted to evaluate the genetic diversity through quantitative and qualitative morphological characteristics in cassava cultivars used by farmers in western region of Paraná State. For both 24 cassava genotypes were collected, and they are subjected to experimental testing conducted in the municipality of Marechal Candido Rondon in the agricultural year 2012/2013. The field trial was implemented following the randomized complete block design with three replications in area situated 24 33 ' south latitude and 54 31' west longitude, and altitude of 420 m. The genetic diversity was performed with the qualitative characters (21 descriptors) and then quantitative (9 descriptors). After tabulating the data was performed obtaining the Euclidean distance matrix for the qualitative and quantitative characteristics for the Mahalanobis and then made up the clustering structure by Tocher methods, UPGMA and nearest neighbor for each array away, with the data quality characteristics also employed the method of Tocher optimization. For qualitative characteristics was found that genotypes Fécula Branca, Cascuda 6, Baianinha 2 and 3 were different from other genotypes of their varieties, existing variability among these varieties or specimens, these examples do not correspond to their respective manifolds. Genotypes Vermelha Uma Rama 1 and 2 were split in all methods, demonstrating that they have inter-variability, evaluated for quality characteristics, not being from the same genetic material. With the data of quantitative traits was found that the genotype of Fécula Branca 3 was different from the other genotypes of the same nomenclature. The other genotypes fécula branca presented in the same group, along with the genotypes of husky. Genotypes Vermelha Uma Rama 1 and 2 were split in all methods, demonstrating that they possess variability among themselves, the set of traits. The Baianinha 1 genotype was different from Baianinha genotypes 2 and 3, showing the existence of variation between them for traits / O presente trabalho foi conduzido para avaliar a diversidade genética através de características morfológicas quantitativas e qualitativas em cultivares de mandioca utilizadas por agricultores da região Oeste do Estado do Paraná. Para tanto foram coletados 24 genótipos de mandioca, sendo os mesmos submetidos a um ensaio experimental conduzido no município de Marechal Cândido Rondon, no ano agrícola 2012/2013. O ensaio de campo foi implantado seguindo o delineamento de blocos ao acaso com três repetições em área situada a 24o 33 de latitude Sul e 54o 31 de longitude Oeste, tendo altitude média de 420 m. A avaliação da diversidade genética foi realizada com os caracteres qualitativos (21 descritores) e depois com os quantitativos (9 descritores). Após a tabulação dos dados realizou-se a obtenção da matriz de distância euclidiana para as características qualitativas e de Mahalanobis para os quantitativos e, em seguida, realizou-se o agrupamento dos genótipos pelos métodos de Tocher, UPGMA e Vizinho mais próximo para cada matriz de distância, com os dados das características qualitativas também empregou-se o método de Otimização de Tocher. Pelas características qualitativas se verificou que os genótipos de Fécula Branca 6, Cascuda 3 e Baianinha 2 foram diferentes dos demais genótipos de suas variedades, existindo variabilidade dentre os exemplares destas variedades ou ainda, estes não correspondem a exemplares de suas respectivas variedades. Os genótipos Vermelha Uma Rama 1 e 2 foram agrupados separadamente em todos os métodos, evidenciando que os mesmos possuem variabilidade entre si, para as características qualitativas avaliadas, não sendo pertencentes ao mesmo material genético. Com os dados das características quantitativas se verificou que o genótipo de Fécula Branca 3, foi diferente dos demais genótipos de mesma nomenclatura. Os demais genótipos de fécula branca se apresentaram no mesmo grupo, juntamente aos genótipos de cascuda. Os genótipos Vermelha Uma Rama 1 e 2 foram agrupados separadamente em todos os métodos, evidenciando que os mesmos possuem variabilidade entre si, para o conjunto de características avaliadas. O genótipo Baianinha 1 foi diferente dos genótipos Baianinha 2 e 3, evidenciando a existência de variação entre os mesmos para as características avaliadas

Classification approaches are used in many areas in order to identify or estimate classes, which different observations belong to. The classification approach, Mahalanobis Taguchi System (MTS) is analyzed and further improved for multi-class classification problems under the scope of this thesis study. MTS tries to explore significant variables and classify a new observation based on its Mahalanobis distance (MD). In this study, first, sample size problems, which are encountered mostly in small data sets, and multicollinearity problems, which constitute some limitations of MTS, are analyzed and a re-sampling approach is explored as a solution. Our re-sampling approach, which only works for data sets with two classes, is a combination of over-sampling and under-sampling. Over-sampling is based on SMOTE, which generates the synthetic observations between the nearest neighbors of observations in the minority class. In addition, MTS models are used to test the performance of several re-sampling parameters, for which the most appropriate values are sought specific to each case. In the second part, multi-class classification methods with MTS are developed. An algorithm, namely Feature Weighted Multi-class MTS-I (FWMMTS-I), is inspired by the descent feature weighted MD. It relaxes adding up of the MDs for variables equally. This provides representations of noisy variables with weights close to zero so that they do not mask the other variables. As a second multi-class classification algorithm, the original MTS method is extended to multi-class problems, which is called Multi-class MTS (MMTS). In addition, a comparable approach to that of Su and Hsiao (2009), which also considers weights of variables, is studied with a modification in MD calculation. It is named as Feature Weighted Multi-class MTS-II (FWMMTS-II). The methods are compared on eight different multi-class data sets using a 5-fold stratified cross validation approach. Results show that FWMMTS-I is as accurate as MMTS, and they are better than FWMMTS-II. Interestingly, the Mahalanobis Distance Classifier (MDC) using all the variables directly in the classification model has performed equally well on the studied data sets.

Submitted by (ana.araujo@ufrpe.br) on 2016-08-03T17:35:12Z No. of bitstreams: 1 Macio Augusto de Albuquerque.pdf: 1005283 bytes, checksum: b9e55eee4b0b853629358e6b2158ba81 (MD5) / Made available in DSpace on 2016-08-03T17:35:12Z (GMT). No. of bitstreams: 1 Macio Augusto de Albuquerque.pdf: 1005283 bytes, checksum: b9e55eee4b0b853629358e6b2158ba81 (MD5) Previous issue date: 2005-02-23 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / The main objective of this research was to propose a systematic to the study and interpretation of the stability of methods in cluster analysis through many cluster algorithms in vegetation data. The data set used came from a survey in the Silviculture Forest at Federal University of Viçosa – MG. To perform the cluster analysis the matrices of Mahalanobis distance were estimated based on the original data and by “bootstrap” resampling. Also the methods of single linkageage, complete linkageage, the average of the distances, the centroid, the medium and the Ward were used. For the detection of the association among the methods it was applied the chi-square test. For the various methods of clustering it was obtained a cofenetical correlation. The results of the associations of methods were very similar, indicating, in principle, that any algorithm of cluster studied is stabilized and exist, in fact, groups among the individuals analyzed. However, it was concluded that themethods coincide with themselves, except the methods of centroid and Ward. Also the centroid methods and average when compared to the Ward, respectively, based on the matrices of Mahalanobis starting from the original data set and “bootstrap”. The methodology proposed is promising to the study and interpretation of the stabilityof methods concerning the cluster analysis in vegetation data. / Objetivou-se propor uma sistemática para o estudo e a interpretação da estabilidade dos métodos em análise de agrupamento, através de vários algoritmos de agrupamento em dados de vegetação. Utilizou-se dados provenientes de um levantamento na Mata da Silvicultura, da Universidade Federal de Viçosa-MG. Para análise de agrupamento foram estimadas as matrizes de distância de Mahalanobis com base nos dados originais e via reamostragem “bootstrap” e aplicados os métodos da ligação simples, ligação completa, médias das distâncias, do centróide, da mediana e do Ward. Para a detecção de associação entre os métodos foi aplicado o teste qui-quadrado. Para os diversos métodos de agrupamento foi obtida a correlação cofenética. Os resultados de associação dos métodos foram semelhantes, indicando em princípio que qualquer algoritmo de agrupamento estudado está estabilizado e existem, de fato, grupos entre os indivíduos observados. No entanto, observou-se que os métodos são coincidentes, exceto osmétodos do centróide e Ward e os métodos do centróide e mediana quando comparados com o de Ward, respectivamente, com base nas matrizes de Mahalanobis a partir dos dados originais e “bootstrap”. A sistemática proposta é promissora para o estudo e a interpretação da estabilidade dos métodos de análise de agrupamento em dados de vegetação.

Orientadores: Victor Hugo Lachos Dávila, Filidor Edilfonso Vilca Labra / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Matemática, Estatística e Computação Científica / Made available in DSpace on 2018-08-17T09:37:18Z (GMT). No. of bitstreams: 1 Carvalho_RignaldoRodrigues_M.pdf: 1849605 bytes, checksum: 07ea5638a2dbfa2227f9a949d4723bbf (MD5) Previous issue date: 2010 / Resumo: O modelo de medição de Barnett é frequentemente usado para comparar vários instrumentos de medição. é comum assumir que os termos aleatórios têm uma distribuição normal. Entretanto, tal suposição faz a inferência vulnerável a observações atípicas por outro lado distribuições de misturas de escala skew-normal tem sido uma interessante alternativa para produzir estimativas robustas tendo a elegância e simplicidade da teoria da máxima verossimilhança. Nós usamos resultados de Lachos et al. (2008) para obter a estimação dos parâmetros via máxima verossimilhança, baseada no algoritmo EM, o qual rende expressões de forma fechada para as equações no passo M. Em seguida desenvolvemos o método de influência local de Zhu e Lee (2001) para avaliar os aspectos de estimação dos parâmetros sob alguns esquemas de perturbação. Os resultados obtidos são aplicados a conjuntos de dados bastante estudados na literatura, ilustrando a utilidade da metodologia proposta / Abstract: The Barnett measurement model is frequently used to comparing several measuring devices. It is common to assume that the random terms have a normal distribution. However, such assumption makes the inference vulnerable to outlying observations whereas scale mixtures of skew-normal distributions have been an interesting alternative to produce robust estimates keeping the elegancy and simplicity of the maximum likelihood theory. We used results in Lachos et al. (2008) for obtaining parameter estimation via maximum likelihood, based on the EM-algorithm, which yields closed form expressions for the equations in the M-step. Then we developed the local influence method to assessing the robustness aspects of these parameter estimates under some usual perturbation schemes. Results obtained for one real data set are reported, illustrating the usefulness of the proposed methodology / Mestrado / Métodos Estatísticos / Mestre em Estatística

In this work, we propose a novel method for a supervised dimensionality reduc- tion, which learns weights of a neural network using an evolutionary algorithm, CMA-ES, optimising the success rate of the k-NN classifier. If no activation func- tions are used in the neural network, the algorithm essentially performs a linear transformation, which can also be used inside of the Mahalanobis distance. There- fore our method can be considered to be a metric learning algorithm. By adding activations to the neural network, the algorithm can learn non-linear transfor- mations as well. We consider reductions to low-dimensional spaces, which are useful for data visualisation, and demonstrate that the resulting projections pro- vide better performance than other dimensionality reduction techniques and also that the visualisations provide better distinctions between the classes in the data thanks to the locality of the k-NN classifier. 1

Intelligent fault detection systems using machine learning can be applied to learn to spot anomalies in signals sampled directly from machinery. As a result, expensive repair costs due to mechanical breakdowns and potential harm to humans due to malfunctioning equipment can be prevented. In recent years, Autoencoders have been applied for fault detection in areas such as industrial manufacturing. It has been shown that they are well suited for the purpose as such models can learn to recognize healthy signals that facilitate the detection of anomalies. The content of this thesis is an investigation into the applicability of Autoencoders for fault detection in mobile robotics by assigning anomaly scores to sampled torque signals based on the Autoencoder reconstruction errors and the Mahalanobis distance to a known distribution of healthy errors. An experiment was carried out by training a model with signals recorded from a four-wheeled mobile robot executing a pre-defined diagnostics routine to stress the motors, and datasets of healthy samples along with three different injected faults were created. The model produced overall greater anomaly scores for one of the fault cases in comparison to the healthy data. However, the two other cases did not yield any difference in anomaly scores due to the faults not impacting the pattern of the signals. Additionally, the Autoencoders ability to isolate a fault to a location was studied by examining the reconstruction errors faulty samples determine whether the errors of signals originating from the faulty component could be used for this purpose. Although we could not confirm this based on the results, fault isolation with Autoencoders could still be possible given more representative signals.

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.

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.