[pt] Nesta tese é proposto um novo método de agrupamento de dados sísmicos para a visualização em mapas sísmicos. Os dados sísmicos (sinal + ruído) têm distribuições assimétricas. A classificação dos dados sísmicos é, atualmente, realizada através de métodos que induzem as referências dos grupos propostos às suas médias. No entanto, a média é sensível aos ruídos e aos outliers e as classificações com este estimador estão sujeitas a distorções nos resultados. Embora outros trabalhos sugiram o uso da mediana nos casos em que as distribuições são assimétricas - devido ao fato deste estimador ser robusto aos ruídos e aos outliers - em nenhum foi encontrado um método que induza as referências dos grupos propostos às medianas no tratamento dos dados sísmicos. O método proposto incluí um algoritmo que induz as referências dos grupos propostos às suas medianas. O tratamento iterativo dos dados sísmicos através da aplicação de uma função não linear adequada ao gradiente descendente gera resultados cujos erros médios quadráticos são inferiores aos dos resultados dos métodos que induzem à média. Um parâmetro existente no algoritmo, a constante de não linearidade, determina a maneira como os dados são induzidos, a partir da média, na direção da mediana. A convergência aos resultados requer poucas iterações no método proposto. O método proposto é uma ferramenta para o dimensionamento de reservatórios de petróleo e serve para a determinação de diferenças entre as propriedades de estruturas geológicas similares. / [en] This thesis suggests the use of a new method of seismic data clustering that can aid in the visualization of seismic maps. Seismic data are primarily made of signal and noise and, due to its dual composition, have asymmetric distributions. Seismic data are traditionally classified by methods that lead the proposed groups` references to their mean values. The mean value is, however, sensitive to noise and outliers and the classification methods that make use of this estimator are, consequently, subjected to generating distorted results. Although other works have suggested the use of the median in cases where the distributions are asymmetric - due to the fact that the estimator is robust with respect to noise and outliers - none have proposed a method that would lead the groups` references to the median while treating seismic data. The method proposed in this work includes, therefore, an algorithm that leads the groups` references to their medians. The iterative treatment of seismic data through the use of a non-linear function that is adequate for the gradient descent generates results with meansquare errors inferior to those of results generated by the use of the mean value. The algorithm`s non- linearity constant determines how the seismic data are led from the mean value towards the median. The proposed method requires little iteration for the results to converge. The proposed method can, therefore, be used as a tool in the sizing of petroleum reservoirs and can also be used to determine the differences between similar geological structures.

As information is becoming more available online and will forevermore be part of any business, the true value of the large amounts of stored data is in the discovery of hidden and unknown relations and connections or traits in the data. The acquisition of these hidden relations can influence strategic decisions which have an impact on the success of a business. Data clustering is one of many methods to partition data into different groups in such a way that data patterns within the same group share some common trait compared to patterns across different groups. This thesis proposes a new artificial immune model for the problem of data clustering. The new model is inspired by the network theory of immunology and differs from its network based predecessor models in its formation of artificial lymphocyte networks. The proposed model is first applied to data clustering problems in stationary environments. Two different techniques are then proposed which enhances the proposed artificial immune model to dynamically determine the number of clusters in a data set with minimal to no user interference. A technique to generate synthetic data sets for data clustering of non-stationary environments is then proposed. Lastly, the original proposed artificial immune model and the enhanced version to dynamically determine the number of clusters are then applied to generated synthetic non-stationary data clustering problems. The influence of the parameters on the clustering performance is investigated for all versions of the proposed artificial immune model and supported by empirical results and statistical hypothesis tests. AFRIKAANS: Soos wat inligting meer aanlyn toeganglik raak en vir altyd meer deel vorm van enige besigheid, is die eintlike waarde van groot hoeveelhede data in die ontdekking van verskuilde en onbekende verwantskappe en konneksies of eienskappe in die data. Die verkryging van sulke verskuilde verwantskappe kan die strategiese besluitneming van ’n besigheid beinvloed, wat weer ’n impak het op die sukses van ’n besigheid. Data groepering is een van baie metodes om data op so ’n manier te groepeer dat data patrone wat deel vorm van dieselfde groep ’n gemeenskaplike eienskap deel in vergelyking met patrone wat verspreid is in ander groepe. Hierdie tesis stel ’n nuwe kunsmatige immuun model voor vir die probleem van data groepering. Die nuwe model is geinspireer deur die netwerk teorie in immunologie en verskil van vorige netwerk gebaseerde modelle deur die model se formasie van kunsmatige limfosiet netwerke. Die voorgestelde model word eers toegepas op data groeperingsprobleme in statiese omgewings. Twee verskillende tegnieke word dan voorgestel wat die voorgestelde kunsmatige immuun model op so ’n manier verbeter dat die model die aantal groepe in ’n data stel dinamies kan bepaal met minimum tot geen gebruiker invloed. ’n Tegniek om kunsmatige data stelle te genereer vir data groepering in dinamiese omgewings word dan voorgestel. Laastens word die oorspronklik voorgestelde model sowel as die verbeterde model wat dinamies die aantal groepe in ’n data stel kan bepaal toegepas op kunsmatig genereerde dinamiese data groeperingsprobleme. Die invloed van die parameters op die groepering prestasie is ondersoek vir alle weergawes van die voorgestelde kunsmatige immuun model en word toegelig deur empiriese resultate en statistiese hipotese toetse. / Thesis (PhD)--University of Pretoria, 2011. / Computer Science / unrestricted

The aim of this master's thesis was to get acquainted with cluster analysis, clustering methods and their theoretical properties. It was necessary select clustering algorithms whose properties will be analyzed, find and select data sets on which these algorithms will be triggered. Also, the goal was to design and implement an application that will evaluate and display clustering results in an appropriate manner. The last step was to analyze the results and compare them with theoretical assumptions.

Artificial Intelligence in Automotive Industry / Cost reduction is one of the main concerns in industry. Companies invest considerably for better performance in end-of-line fault diagnosis systems. A common strategy is to use data obtained from existing instrumentation. This research investigates the challenge of learning from historical data that have already been collected by companies. Machine learning is basically one of the most common and powerful techniques of artificial intelligence that can learn from data and identify fault features with no need for human interaction. In this research, labeled sound and vibration measurements are processed into fault signatures for vehicle starter motors and alternators. A fault detection and identification system has been developed to identify fault types for end-of-line testing of motors. However, labels are relatively difficult to obtain, expensive, time consuming and require experienced humans, while unlabeled samples needs less effort to collect. Thus, learning from unlabeled data together with the guidance of few labels would be a better solution. Furthermore, in this research, learning from unlabeled data with absolutely no human intervention is also implemented and discussed as well. / Thesis / Master of Applied Science (MASc)

Machine learning is evidenced as a research area with the main purpose of developing computational methods that are capable of learning with their previously acquired experiences. Although a large amount of machine learning techniques has been proposed and successfully applied in real systems, there are still many challenging issues, which need be addressed. In the last years, an increasing interest in techniques based on complex networks (large-scale graphs with nontrivial connection patterns) has been verified. This emergence is explained by the inherent advantages provided by the complex network representation, which is able to capture the spatial, topological and functional relations of the data. In this work, we investigate the new features and possible advantages offered by complex networks in the machine learning domain. In fact, we do show that the network-based approach really brings interesting features for supervised, semisupervised, and unsupervised learning. Specifically, we reformulate a previously proposed particle competition technique for both unsupervised and semisupervised learning using a stochastic nonlinear dynamical system. Moreover, an analytical analysis is supplied, which enables one to predict the behavior of the proposed technique. In addition to that, data reliability issues are explored in semisupervised learning. Such matter has practical importance and is found to be of little investigation in the literature. With the goal of validating these techniques for solving real problems, simulations on broadly accepted databases are conducted. Still in this work, we propose a hybrid supervised classification technique that combines both low and high orders of learning. The low level term can be implemented by any classification technique, while the high level term is realized by the extraction of features of the underlying network constructed from the input data. Thus, the former classifies the test instances by their physical features, while the latter measures the compliance of the test instances with the pattern formation of the data. Our study shows that the proposed technique not only can realize classification according to the semantic meaning of the data, but also is able to improve the performance of traditional classification techniques. Finally, it is expected that this study will contribute, in a relevant manner, to the machine learning area / Aprendizado de máquina figura-se como uma área de pesquisa que visa a desenvolver métodos computacionais capazes de aprender com a experiência. Embora uma grande quantidade de técnicas de aprendizado de máquina foi proposta e aplicada, com sucesso, em sistemas reais, existem ainda inúmeros problemas desafiantes que necessitam ser explorados. Nos últimos anos, um crescente interesse em técnicas baseadas em redes complexas (grafos de larga escala com padrões de conexão não triviais) foi verificado. Essa emergência é explicada pelas inerentes vantagens que a representação em redes complexas traz, sendo capazes de capturar as relações espaciais, topológicas e funcionais dos dados. Nesta tese, serão investigadas as possíveis vantagens oferecidas por redes complexas quando utilizadas no domínio de aprendizado de máquina. De fato, será mostrado que a abordagem por redes realmente proporciona melhorias nos aprendizados supervisionado, semissupervisionado e não supervisionado. Especificamente, será reformulada uma técnica de competição de partículas para o aprendizado não supervisionado e semissupervisionado por meio da utilização de um sistema dinâmico estocástico não linear. Em complemento, uma análise analítica de tal modelo será desenvolvida, permitindo o entendimento evolucional do modelo no tempo. Além disso, a questão de confiabilidade de dados será investigada no aprendizado semissupervisionado. Tal tópico tem importância prática e é pouco estudado na literatura. Com o objetivo de validar essas técnicas em problemas reais, simulações computacionais em bases de dados consagradas pela literatura serão conduzidas. Ainda nesse trabalho, será proposta uma técnica híbrica de classificação supervisionada que combina tanto o aprendizado de baixo como de alto nível. O termo de baixo nível pode ser implementado por qualquer técnica de classificação tradicional, enquanto que o termo de alto nível é realizado pela extração das características de uma rede construída a partir dos dados de entrada. Nesse contexto, aquele classifica as instâncias de teste segundo qualidades físicas, enquanto que esse estima a conformidade da instância de teste com a formação de padrões dos dados. Os estudos aqui desenvolvidos mostram que o método proposto pode melhorar o desempenho de técnicas tradicionais de classificação, além de permitir uma classificação de acordo com o significado semântico dos dados. Enfim, acredita-se que este estudo possa gerar contribuições relevantes para a área de aprendizado de máquina.

La radiothérapie est une des méthodes principales utilisée dans le traitement thérapeutique des tumeurs malignes. Pour améliorer son efficacité, deux problèmes essentiels doivent être soigneusement traités : la prédication fiable des résultats thérapeutiques et la segmentation précise des volumes tumoraux. La tomographie d’émission de positrons au traceur Fluoro- 18-déoxy-glucose (FDG-TEP) peut fournir de manière non invasive des informations significatives sur les activités fonctionnelles des cellules tumorales. Les objectifs de cette thèse sont de proposer: 1) des systèmes fiables pour prédire les résultats du traitement contre le cancer en utilisant principalement des caractéristiques extraites des images FDG-TEP; 2) des algorithmes automatiques pour la segmentation de tumeurs de manière précise en TEP et TEP-TDM. La théorie des fonctions de croyance est choisie dans notre étude pour modéliser et raisonner des connaissances incertaines et imprécises pour des images TEP qui sont bruitées et floues. Dans le cadre des fonctions de croyance, nous proposons une méthode de sélection de caractéristiques de manière parcimonieuse et une méthode d’apprentissage de métriques permettant de rendre les classes bien séparées dans l’espace caractéristique afin d’améliorer la précision de classification du classificateur EK-NN. Basées sur ces deux études théoriques, un système robuste de prédiction est proposé, dans lequel le problème d’apprentissage pour des données de petite taille et déséquilibrées est traité de manière efficace. Pour segmenter automatiquement les tumeurs en TEP, une méthode 3-D non supervisée basée sur le regroupement évidentiel (evidential clustering) et l’information spatiale est proposée. Cette méthode de segmentation mono-modalité est ensuite étendue à la co-segmentation dans des images TEP-TDM, en considérant que ces deux modalités distinctes contiennent des informations complémentaires pour améliorer la précision. Toutes les méthodes proposées ont été testées sur des données cliniques, montrant leurs meilleures performances par rapport aux méthodes de l’état de l’art. / Radiation therapy is one of the most principal options used in the treatment of malignant tumors. To enhance its effectiveness, two critical issues should be carefully dealt with, i.e., reliably predicting therapy outcomes to adapt undergoing treatment planning for individual patients, and accurately segmenting tumor volumes to maximize radiation delivery in tumor tissues while minimize side effects in adjacent organs at risk. Positron emission tomography with radioactive tracer fluorine-18 fluorodeoxyglucose (FDG-PET) can noninvasively provide significant information of the functional activities of tumor cells. In this thesis, the goal of our study consists of two parts: 1) to propose reliable therapy outcome prediction system using primarily features extracted from FDG-PET images; 2) to propose automatic and accurate algorithms for tumor segmentation in PET and PET-CT images. The theory of belief functions is adopted in our study to model and reason with uncertain and imprecise knowledge quantified from noisy and blurring PET images. In the framework of belief functions, a sparse feature selection method and a low-rank metric learning method are proposed to improve the classification accuracy of the evidential K-nearest neighbor classifier learnt by high-dimensional data that contain unreliable features. Based on the above two theoretical studies, a robust prediction system is then proposed, in which the small-sized and imbalanced nature of clinical data is effectively tackled. To automatically delineate tumors in PET images, an unsupervised 3-D segmentation based on evidential clustering using the theory of belief functions and spatial information is proposed. This mono-modality segmentation method is then extended to co-segment tumor in PET-CT images, considering that these two distinct modalities contain complementary information to further improve the accuracy. All proposed methods have been performed on clinical data, giving better results comparing to the state of the art ones.

Machine learning is evidenced as a research area with the main purpose of developing computational methods that are capable of learning with their previously acquired experiences. Although a large amount of machine learning techniques has been proposed and successfully applied in real systems, there are still many challenging issues, which need be addressed. In the last years, an increasing interest in techniques based on complex networks (large-scale graphs with nontrivial connection patterns) has been verified. This emergence is explained by the inherent advantages provided by the complex network representation, which is able to capture the spatial, topological and functional relations of the data. In this work, we investigate the new features and possible advantages offered by complex networks in the machine learning domain. In fact, we do show that the network-based approach really brings interesting features for supervised, semisupervised, and unsupervised learning. Specifically, we reformulate a previously proposed particle competition technique for both unsupervised and semisupervised learning using a stochastic nonlinear dynamical system. Moreover, an analytical analysis is supplied, which enables one to predict the behavior of the proposed technique. In addition to that, data reliability issues are explored in semisupervised learning. Such matter has practical importance and is found to be of little investigation in the literature. With the goal of validating these techniques for solving real problems, simulations on broadly accepted databases are conducted. Still in this work, we propose a hybrid supervised classification technique that combines both low and high orders of learning. The low level term can be implemented by any classification technique, while the high level term is realized by the extraction of features of the underlying network constructed from the input data. Thus, the former classifies the test instances by their physical features, while the latter measures the compliance of the test instances with the pattern formation of the data. Our study shows that the proposed technique not only can realize classification according to the semantic meaning of the data, but also is able to improve the performance of traditional classification techniques. Finally, it is expected that this study will contribute, in a relevant manner, to the machine learning area / Aprendizado de máquina figura-se como uma área de pesquisa que visa a desenvolver métodos computacionais capazes de aprender com a experiência. Embora uma grande quantidade de técnicas de aprendizado de máquina foi proposta e aplicada, com sucesso, em sistemas reais, existem ainda inúmeros problemas desafiantes que necessitam ser explorados. Nos últimos anos, um crescente interesse em técnicas baseadas em redes complexas (grafos de larga escala com padrões de conexão não triviais) foi verificado. Essa emergência é explicada pelas inerentes vantagens que a representação em redes complexas traz, sendo capazes de capturar as relações espaciais, topológicas e funcionais dos dados. Nesta tese, serão investigadas as possíveis vantagens oferecidas por redes complexas quando utilizadas no domínio de aprendizado de máquina. De fato, será mostrado que a abordagem por redes realmente proporciona melhorias nos aprendizados supervisionado, semissupervisionado e não supervisionado. Especificamente, será reformulada uma técnica de competição de partículas para o aprendizado não supervisionado e semissupervisionado por meio da utilização de um sistema dinâmico estocástico não linear. Em complemento, uma análise analítica de tal modelo será desenvolvida, permitindo o entendimento evolucional do modelo no tempo. Além disso, a questão de confiabilidade de dados será investigada no aprendizado semissupervisionado. Tal tópico tem importância prática e é pouco estudado na literatura. Com o objetivo de validar essas técnicas em problemas reais, simulações computacionais em bases de dados consagradas pela literatura serão conduzidas. Ainda nesse trabalho, será proposta uma técnica híbrica de classificação supervisionada que combina tanto o aprendizado de baixo como de alto nível. O termo de baixo nível pode ser implementado por qualquer técnica de classificação tradicional, enquanto que o termo de alto nível é realizado pela extração das características de uma rede construída a partir dos dados de entrada. Nesse contexto, aquele classifica as instâncias de teste segundo qualidades físicas, enquanto que esse estima a conformidade da instância de teste com a formação de padrões dos dados. Os estudos aqui desenvolvidos mostram que o método proposto pode melhorar o desempenho de técnicas tradicionais de classificação, além de permitir uma classificação de acordo com o significado semântico dos dados. Enfim, acredita-se que este estudo possa gerar contribuições relevantes para a área de aprendizado de máquina.