[pt] A medição e classificação de descargas parciais constituem uma importante ferramenta de avaliação dos sistemas de isolamento utilizados em equipamentos de alta tensão. Após o pré-processamento dos dados, que captura, digitaliza e filtra o sinal de descargas parciais, geralmente eliminando os ruídos, existem basicamente duas etapas principais, que são a extração de características e a classificação de padrões. As descargas parciais contêm um conjunto de características discriminatórias únicas que lhes permitem ser reconhecidas. Assim, o primeiro procedimento no processo de classificação é definir quais delas podem ser utilizadas e qual o método de extração destas características. O fenômeno de descargas parciais tem uma natureza transitória e é caracterizado por correntes pulsantes com uma duração de vários nanossegundos até poucos microssegundos. Sua magnitude não é sempre proporcional ao dano causado, sendo que descargas de pequena magnitude podem levar rapidamente à evolução de um defeito. Por isso a necessidade de se entender bem este fenômeno e saber interpretar os dados. Além disso, equipamentos de alta tensão de grande porte, como motores e geradores, podem apresentar mais de uma fonte interna de descargas parciais, sendo importante separar os sinais dessas diferentes fontes antes de realizar a classificação. No caso de outros equipamentos de alta tensão de menor porte, como para-raios e transformadores de corrente de subestação, a simples detecção da presença de descargas parciais interna ao equipamento, independente do número de fontes, já é suficiente para indicar a retirada de operação destes equipamentos, dado seu baixo custo relativo e o elevado grau de importância destes para a confiabilidade do sistema onde estão inseridos. Para um diagnóstico completo e confíável de isolamentos de alta tensão, há a demanda por um sistema de análise capaz de promover com eficácia a detecção de descargas parciais internas aos equipamentos, a separação das diversas fontes de descargas parciais, no caso dos equipamentos de grande porte, bem como realizar a correta classificação do tipo de defeito, com base principalmente na análise das características discriminantes das diferentes fontes e na assinatura dos sinais para os diferentes defeitos. Este estudo contribui para o preenchimento desta lacuna, apresentando metodologias que se mostram robustas e precisas nos testes realizados, de modo que possam efetivamente orientar os especialistas em manutenção na tomada de decisões. Para fazer isso, são propostas novas variáveis capazes de extrair informações relevantes de sinais no tempo medidos em diversos tipos de isolamentos, sendo aplicadas aqui em dados obtidos em campo e em laboratório para avaliar sua eficácia na tarefa. Essas informações são tratadas utilizando técnicas de classificação de padrões e inteligência artificial para determinar de forma automática a presença de descargas parciais, o número de fontes diferentes e o tipo de defeito nos isolamentos de alta tensão utilizados no estudo. Outra contribuição do estudo é a criação de um banco de dados histórico, baseada em processamento de imagem, com padrões de mapas de descargas parciais conhecidos na literatura em máquinas rotativas, para serem utilizados na classificação de novos mapas medidos neste tipo de equipamento. / [en] Measurement and classification of partial discharges are an important tool for the evaluation of insulation systems used in high voltage equipments. After pre-processing of data, which captures, scans and filters the signal of partial discharges, generally eliminating noises, there are basically two main steps, which are the extraction of characteristics and the pattern classification. Partial discharges contain a set of unique discriminatory characteristics that allow them to be recognized. Thus, the first procedure in the classification process is to define which of them can be used and which is the method for extraction of those characteristics. The phenomenon of partial discharges has a transient nature and is characterized by pulsating currents with a duration of several nanoseconds up to a few microseconds. Its magnitude is not always proportional to the damage caused, and discharges of small magnitude can quickly lead to the evolution of a failure. Therefore the need to understand this phenomenon well and to know how to interpret the data. In addition, large high voltage equipments such as motors and generators may have more than one internal source of partial discharges, and it is important to separate the signals from those different sources prior to classification. In the case of smaller high voltage equipments, as surge arrester and substation current transformers, the simple detection of the presence of partial discharges inside the equipment, regardless of the number of sources, is sufficient to indicate the withdrawal of operation of the equipment, given their low relative cost and the high degree of importance of these to the reliability of the system where they are part of. For a complete and reliable diagnosis of high voltage insulations, there is a demand for an analysis system capable of effectively promoting the detection of the partial discharges internal to the equipments, the separation of the various sources of partial discharges in the case of large equipments, as well as to carry out the correct classification of the type of failure. The system should be based mainly on the analysis of the discriminating characteristics of the different sources and the signature of the signals for the different failure. This study contributes to fill this gap by presenting methodologies that are robust and accurate in the tests performed, so that they can effectively guide maintenance specialists in decision making. To do this, new variables are proposed to extract relevant information from time signals measured in various types of insulations, being applied here in field and laboratory data to evaluate their effectiveness in the task. This information is treated using standard classification techniques and artificial intelligence to automatically determine the presence of partial discharges, the number of different sources and the type of defect in the high voltage insulations used in the study. Another contribution of the study is the creation of a historical database, based on image processing, with partial discharge map patterns known in the literature on rotating machines, to be used in the classification of new maps measured in this type of equipment.

本論文基於具備捕捉影像深度的RGBD影音串流裝置－Kinect感測器，在簡報場域中，作為擷取簡報者肢體動作、表情、以及語言表現模式的設備。首先我們提出在特定時段內的表現模式，可以經由大眾的評估，而具有喜歡／不喜歡的性質，我們將其分別命名為Period of Like(POL)以及Period of Dislike(POD)。論文中並以三種Kinect SDK所提供的影像特徵：動畫單元、骨架關節點、以及3D臉部頂點，輔以35位評估者所提供之評估資料，以POD/POL取出的特徵模式，分析是否具有一致性，以及是否可用於未來預測。最後將研究結果開發應用於原型程式，期許這樣的預測系統，能夠為在簡報中表現不佳而困擾的人們，提點其優劣之處，以作為後續改善之依據。 / In this thesis, we capture body movements, facial expressions, and voice data of subjects in the presentation scenario using RGBD-capable Kinect sensor. The acquired videos were accessed by a group of reviewers to indicate their preferences/aversions to the presentation style. We denote the two classes of ruling as Period of Like (POL) and Period of Dislike (POD), respectively. We then employ three types of image features, namely, animation units (AU), skeletal joints, and 3D face vertices to analyze the consistency of the evaluation result, as well as the ability to classify unseen footage based on the training data supplied by 35 evaluators. Finally, we develop a prototype program to help users to identify their strength/weakness during their presentation so that they can improve their skills accordingly.

Made available in DSpace on 2016-06-02T19:06:12Z (GMT). No. of bitstreams: 1 5931.pdf: 2655791 bytes, checksum: 6eafe016c175143a8d55692b4681adfe (MD5) Previous issue date: 2014-05-05 / Financiadora de Estudos e Projetos / Classification is a very important step in pattern recognition, as it aims to categorize objects from a set of inherent features, through its labeling. This process can be supervised, when there is a sample set of labeled training classes, semi-supervised, when the number of labeled samples is limited or nearly inexistent, or unsupervised, where there are no labeled samples. This project proposes to explore minimum information paths in graphs for classification problems, through the definition of a supervised, non-parametric, graph-based classification method, by means of a contextual approach. This method proposes to construct a graph from a set of training samples, where the samples are represented by vertices and the edges are links between samples that belongs to a neighborhood system. From the graph construction, the method calculates the local observed Fisher information, a measurement based on the Potts model, for all vertices, identifying the amount of information that each sample has. Generally, different class vertices when connected by an edge, have a high information level. After that, it is necessary to weight the edges by means of a function that penalizes connecting vertices with high information. During this process, it is possible to identify and select high information vertices, which will be chosen to be prototype vertices, namely, the nodes that define the classes boundaries. After the definition, the method proposes that each prototype sample conquer the remaining samples by offering the shortest path in terms of information, so that when a sample is conquered it receives the label of the winning prototype, occurring the classification. To evaluate the proposed method, statistical methods to estimate the error rates, such as Hold-out, K-fold and Leave-One- Out Cross-Validation will be considered. The obtained results indicate that the method can be a viable alternative to the existing classification techniques. / A classificação é uma etapa muito importante em reconhecimento de padrões, pois ela tem o objetivo de categorizar objetos a partir de um conjunto de características inerentes a ele, atribuindo-lhe um rótulo. Esse processo de classificação pode ser supervisionado, quando existe um conjunto de amostras de treinamento rotuladas que representam satisfatoriamente as classes, semi-supervisionado, quando o conjunto de amostras é limitado ou quase inexistente, ou não-supervisionado, quando não existem amostras rotuladas. Este trabalho propõe explorar caminhos de mínima informação em grafos para problemas de classificação, por meio da criação de um método de classificação supervisionado, não paramétrico, baseado em grafos, seguindo uma abordagem contextual. Esse método propõe a construção de um grafo a partir do conjunto de amostras de treinamento, onde as amostras serão representadas pelos vértices e as arestas serão as ligações entre amostras pertencentes a uma relação de adjacência. A partir da construção do grafo o método faz o calculo da informação de Fisher Local Observada, uma medida baseada no modelo de Potts, para todos os vértices, identificando o grau de informação que cada um possui. Geralmente vértices de classes distintas quando conectados por uma aresta possuem alta informação (bordas). Feito o calculo da informação, é necessário ponderar as arestas por meio de uma função que penaliza a ligação de vértices com alta informação. Enquanto as arestas são ponderadas é possível identificar e selecionar vértices altamente informativos os quais serão escolhidos para serem vértices protótipos, ou seja, os vértices que definem a região de borda. Depois de ponderadas as arestas e definidos os protótipos, o método propõe que cada protótipo conquiste as amostras oferecendo o menor caminho até ele, de modo que quando uma amostra é conquistada ela receba o rótulo do protótipo que a conquistou, ocorrendo a classificação. Para avaliar o método serão utilizados métodos estatísticos para estimar as taxas de acertos, como K-fold, Hold-out e Leave-one-out Cross- Validation. Os resultados obtidos indicam que o método pode ser um uma alternativa viável as técnicas de classificação existentes.

Made available in DSpace on 2014-12-17T14:56:07Z (GMT). No. of bitstreams: 1 NaiyanHCL_DISSERT.pdf: 1452285 bytes, checksum: 018fb1e8fa51e8f7094cce68a18c6c73 (MD5) Previous issue date: 2012-08-13 / Reinforcement learning is a machine learning technique that, although finding a large number of applications, maybe is yet to reach its full potential. One of the inadequately tested possibilities is the use of reinforcement learning in combination with other methods for the solution of pattern classification problems. It is well documented in the literature the problems that support vector machine ensembles face in terms of generalization capacity. Algorithms such as Adaboost do not deal appropriately with the imbalances that arise in those situations. Several alternatives have been proposed, with varying degrees of success. This dissertation presents a new approach to building committees of support vector machines. The presented algorithm combines Adaboost algorithm with a layer of reinforcement learning to adjust committee parameters in order to avoid that imbalances on the committee components affect the generalization performance of the final hypothesis. Comparisons were made with ensembles using and not using the reinforcement learning layer, testing benchmark data sets widely known in area of pattern classification / A aprendizagem por refor?o ? uma t?cnica de aprendizado de m?quina que, embora j? tenha encontrado uma grande quantidade de aplica??es, talvez ainda n?o tenha alcan?ado seu pleno potencial. Uma das possibilidades que n?o foi devidamente testada at? hoje foi a utiliza??o da aprendizagem por refor?o em conjunto com outros m?todos para a solu??o de problemas de classifica??o de padr?es. ? bem documentada na literatura a problem?tica que ensembles de m?quinas de vetor de suporte encontram em termos de capacidade de generaliza??o. Algoritmos como Adaboost n?o lidam apropriadamente com os desequil?brios que podem surgir nessas situa??es. V?rias alternativas j? foram propostas, com margens variadas de sucesso. Esta disserta??o apresenta uma nova abordagem para a constru??o de comit?s de m?quinas de vetor de suporte. O algoritmo apresentado combina o algoritmo Adaboost com uma camada de aprendizagem por refor?o, para ajustar par?metros do comit? evitando que desequil?brios nos classificadores componentes do comit? prejudiquem o desempenho de generaliza??o da hip?tese final. Foram efetuadas compara??es de comit?s com e sem essa camada adicional de aprendizagem por refor?o, testando conjuntos de dados benchmarks amplamente conhecidos na ?rea de classifica??o de padr?es

Made available in DSpace on 2014-12-17T14:56:13Z (GMT). No. of bitstreams: 1 CarlosAAP_DISSERT.pdf: 1150903 bytes, checksum: a90e625336bbabe7e96da74cb85ee7aa (MD5) Previous issue date: 2013-01-31 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior / The pattern classification is one of the machine learning subareas that has the most outstanding. Among the various approaches to solve pattern classification problems, the Support Vector Machines (SVM) receive great emphasis, due to its ease of use and good generalization performance. The Least Squares formulation of SVM (LS-SVM) finds the solution by solving a set of linear equations instead of quadratic programming implemented in SVM. The LS-SVMs provide some free parameters that have to be correctly chosen to achieve satisfactory results in a given task. Despite the LS-SVMs having high performance, lots of tools have been developed to improve them, mainly the development of new classifying methods and the employment of ensembles, in other words, a combination of several classifiers. In this work, our proposal is to use an ensemble and a Genetic Algorithm (GA), search algorithm based on the evolution of species, to enhance the LSSVM classification. In the construction of this ensemble, we use a random selection of attributes of the original problem, which it splits the original problem into smaller ones where each classifier will act. So, we apply a genetic algorithm to find effective values of the LS-SVM parameters and also to find a weight vector, measuring the importance of each machine in the final classification. Finally, the final classification is obtained by a linear combination of the decision values of the LS-SVMs with the weight vector. We used several classification problems, taken as benchmarks to evaluate the performance of the algorithm and compared the results with other classifiers / A classifica??o de padr?es ? uma das sub?reas do aprendizado de m?quina que possui maior destaque. Entre as v?rias t?cnicas para resolver problemas de classifica??o de padr?es, as M?quinas de Vetor de Suporte (do ingl?s, Support Vector Machines ou SVM) recebem grande ?nfase, devido a sua facilidade de uso e boa capacidade de generaliza??o. A formula??o por M?nimos Quadrados da SVM (do ingl?s, Least Squares Support Vector Machines ou LS-SVM) encontra um hiperplano de separa??o ?tima atrav?s da solu??o de um sistema de equa??es lineares, evitando assim o uso da programa??o quadr?tica implementada na SVM. As LS-SVMs fornecem alguns par?metros livres que precisam ser corretamente selecionados para alcan?ar resultados satisfat?rios em uma determinada tarefa. Apesar das LS-SVMs possuir elevado desempenho, v?rias ferramentas tem sido desenvolvidas para aprimor?-la, principalmente o desenvolvimento de novos m?todos de classifica??o e a utiliza??o de comit?s de m?quinas, ou seja, a combina??o de v?rios classificadores. Neste trabalho, n?s propomos tanto o uso de um comit? de m?quinas quanto o uso de um Algoritmo Gen?tico (AG), algoritmo de busca baseada na evolu??o das esp?cies, para aprimorar o poder de classifica??o da LS-SVM. Na constru??o desse comit?, utilizamos uma sele??o aleat?ria de atributos do problema original, que divide o problema original em outros menores onde cada classificador do comit? vai atuar. Ent?o, aplicamos o AG para encontrar valores efetivos para os par?metros de cada LS-SVM e tamb?m encontrando um vetor de pesos, medindo a import?ncia de cada m?quina na classifica??o final. Por fim, a classifica??o final ? dada por uma combina??o linear das respostas de cada m?quina ponderadas pelos pesos. Foram utilizados v?rios problemas de classifica??o, tidos como benchmarks, para avaliar o desempenho do algoritmo e comparamos os resultados obtidos com outros classificadores

No description available.

Geospatial analysis involves application of statistical methods, algorithms and information retrieval techniques to geospatial data. It incorporates time into spatial databases and facilitates investigation of land cover (LC) dynamics through data, model, and analytics. LC dynamics induced by human and natural processes play a major role in global as well as regional scale patterns, which in turn influence weather and climate. Hence, understanding LC dynamics at the local / regional as well as at global levels is essential to evolve appropriate management strategies to mitigate the impacts of LC changes. This can be captured through the multi-resolution remote sensing (RS) data. However, with the advancements in sensor technologies, suitable algorithms and techniques are required for optimal integration of information from multi-resolution sensors which are cost effective while overcoming the possible data and methodological constraints. In this work, several per-pixel traditional and advanced classification techniques have been evaluated with the multi-resolution data along with the role of ancillary geographical data on the performance of classifiers. Techniques for linear and non-linear un-mixing, endmember variability and determination of spatial distribution of class components within a pixel have been applied and validated on multi-resolution data. Endmember estimation method is proposed and its performance is compared with manual, semi-automatic and fully automatic methods of endmember extraction. A novel technique - Hybrid Bayesian Classifier is developed for per pixel classification where the class prior probabilities are determined by un-mixing a low spatial-high spectral resolution multi-spectral data while posterior probabilities are determined from the training data obtained from ground, that are assigned to every pixel in a high spatial-low spectral resolution multi-spectral data in Bayesian classification. These techniques have been validated with multi-resolution data for various landscapes with varying altitudes. As a case study, spatial metrics and cellular automata based models applied for rapidly urbanising landscape with moderate altitude has been carried out.

種々の分類器を対象として，有限個の学習データのみが利用可能である現実においても理論的に的確で計算量的にも実際的な，分類器性能評価手法を提案する．分類器評価における難しさは，有限データのみの利用に起因する分類誤り推定に伴う偏りの発生にある．この困難を解決するため，「境界曖昧性」と呼ばれる新しい評価尺度を提案し，それを用いる評価法の有用性を，3種の分類器と13個のデータセットを用いた実験を通して実証する． / We propose a general method that makes accurate evaluation of any classifier model for realistic tasks, both in a theoretical sense despite the finiteness of the available data, and in a practical sense in terms of computation costs. The classifier evaluation challenge arises from the bias of the classification error estimate that is only based on finite data. We bypass this existing difficulty by proposing a new classifier evaluation measure called "boundary uncertainty'' whose estimate based on finite data can be considered a reliable representative of its expectation based on infinite data, and demonstrate the potential of our approach on three classifier models and thirteen datasets. / 博士(工学) / Doctor of Philosophy in Engineering / 同志社大学 / Doshisha University

This thesis deals with machine learning techniques for the extraction of structure and the analysis of the vertebrate olfactory pathway based on related methods. Some of its main contributions are summarized below. We have performed a systematic investigation for classification in biomedical images with the goal of recognizing a material in these images by its texture. This investigation included (i) different measures for evaluating the importance of image descriptors (features), (ii) methods to select a feature set based on these evaluations, and (iii) classification algorithms. Image features were evaluated according to their estimated relevance for the classification task and their redundancy with other features. For this purpose, we proposed a framework for relevance and redundancy measures and, within this framework, we proposed two new measures. These were the value difference metric and the fit criterion. Both measures performed well in comparison with other previously used ones for evaluating features. We also proposed a Hopfield network as a method for feature selection, which in experiments gave one of the best results relative to other previously used approaches. We proposed a genetic algorithm for clustering and tested it on several realworld datasets. This genetic algorithm was novel in several ways, including (i) the use of intra-cluster distance as additional optimization criterion, (ii) an annealing procedure, and (iii) adaptation of mutation rates. As opposed to many conventional clustering algorithms, our optimization framework allowed us to use different cluster validation measures including those which do not rely on cluster centroids. We demonstrated the use of the clustering algorithm experimentally with several cluster validity measures as optimization criteria. We compared the performance of our clustering algorithm to that of the often-used fuzzy c-means algorithm on several standard machine learning datasets from the University of California/Urvine (UCI) and obtained good results. The organization of representations in the brain has been observed at several stages of processing to spatially decompose input from the environment into features that are somehow relevant from a behavioral or perceptual standpoint. For the perception of smells, the analysis of such an organization, however, is not as straightforward because of the missing metric. Some studies report spatial clusters for several combinations of physico-chemical properties in the olfactory bulb at the level of the glomeruli. We performed a systematic study of representations based on a dataset of activity-related images comprising more than 350 odorants and covering the whole spatial array of the first synaptic level in the olfactory system. We found clustered representations for several physico-chemical properties. We compared the relevance of these properties to activations and estimated the size of the coding zones. The results confirmed and extended previous studies on olfactory coding for physico-chemical properties. Particularly of interest was the spatial progression by carbon chain that we found. We discussed our estimates of relevance and coding size in the context of processing strategies. We think that the results obtained in this study could guide the search into olfactory coding primitives and the understanding of the stimulus space. In a second study on representations in the olfactory bulb, we grouped odorants together by perceptual categories, such as floral and fruity. By the application of the same statistical methods as in the previous study, we found clustered zones for these categories. Furthermore, we found that distances between spatial representations were related to perceptual differences in humans as reported in the literature. This was possibly the first time that such an analysis had been done. Apart from pointing towards a spatial decomposition by perceptual dimensions, results indicate that distance relationships between representations could be perceptually meaningful. In a third study, we modeled axon convergence from olfactory receptor neurons to the olfactory bulb. Sensory neurons were stimulated by a set of biologically-relevant odors, which were described by a set of physico-chemical properties that covaried with the neural and glomerular population activity in the olfactory bulb. Convergence was mediated by the covariance between olfactory neurons. In our model, we could replicate the formation of glomeruli and concentration coding as reported in the literature, and further, we found that the spatial relationships between representational zones resulting from our model correlated with reported perceptual differences between odor categories. This shows that natural statistics, including similarity of physico-chemical structure of odorants, can give rise to an ordered arrangement of representations at the olfactory bulb level where the distances between representations are perceptually relevant. / <p>QC 20120224</p>

Made available in DSpace on 2014-12-17T15:46:59Z (GMT). No. of bitstreams: 1 LauraEASS_TESE.pdf: 2447411 bytes, checksum: 3e442431965058383423623bc7751de0 (MD5) Previous issue date: 2012-02-02 / Conselho Nacional de Desenvolvimento Cient?fico e Tecnol?gico / Traditional applications of feature selection in areas such as data mining, machine learning and pattern recognition aim to improve the accuracy and to reduce the computational cost of the model. It is done through the removal of redundant, irrelevant or noisy data, finding a representative subset of data that reduces its dimensionality without loss of performance. With the development of research in ensemble of classifiers and the verification that this type of model has better performance than the individual models, if the base classifiers are diverse, comes a new field of application to the research of feature selection. In this new field, it is desired to find diverse subsets of features for the construction of base classifiers for the ensemble systems. This work proposes an approach that maximizes the diversity of the ensembles by selecting subsets of features using a model independent of the learning algorithm and with low computational cost. This is done using bio-inspired metaheuristics with evaluation filter-based criteria / A aplica??o tradicional da sele??o de atributos em diversas ?reas como minera??o de dados, aprendizado de m?quina e reconhecimento de padr?es visa melhorar a acur?cia dos modelos constru?dos com a base de dados, ao retirar dados ruidosos, redundantes ou irrelevantes, e diminuir o custo computacional do modelo, ao encontrar um subconjunto representativo dos dados que diminua sua dimensionalidade sem perda de desempenho. Com o desenvolvimento das pesquisas com comit?s de classificadores e a verifica??o de que esse tipo de modelo possui melhor desempenho que os modelos individuais, dado que os classificadores base sejam diversos, surge uma nova aplica??o ?s pesquisas com sele??o de atributos, que ? a de encontrar subconjuntos diversos de atributos para a constru??o dos classificadores base de comit?s de classificadores. O presente trabalho prop?e uma abordagem que maximiza a diversidade de comit?s de classificadores atrav?s da sele??o de subconjuntos de atributos utilizando um modelo independente do algoritmo de aprendizagem e de baixo custo computacional. Isso ? feito utilizando metaheur?sticas bioinspiradas com crit?rios de avalia??o baseados em filtro