Thesis (Ph.D.) -- Mississippi State University. Department of Electrical and Computer Engineering. / Title from title screen. Includes bibliographical references.

Thesis (Ph.D.)--Mississippi State University. Department of Electrical and Computer Engineering. / Title from title screen. Includes bibliographical references.

Τα συστήματα διεπαφών ανθρώπινου εγκεφάλου-υπολογιστή (BCIs: Brain-Computer Interfaces) απαιτούν την πραγματικού χρόνου, αποτελεσματική επεξεργασία των μετρήσεων των ηλεκτροεγκεφαλογραφικών (ΗΕΓ) σημάτων του χρήστη τους, προκειμένου να μεταφράσουν τις νοητικές διεργασίες/προθέσεις του σε σήματα ελέγχου εξωτερικών διατάξεων ή συστημάτων. Στο πλαίσιο της εργασίας αυτής μελετήθηκε το θεωρητικό υπόβαθρο του προβλήματος και αναλύθηκαν συνοπτικά οι κυριότερες τεχνικές που χρησιμοποιούνται σήμερα. Επιπρόσθετα, παρουσιάστηκε μία μέθοδος ταξινόμησης των νοητικών προθέσεων της αριστερής και δεξιάς κίνησης των χεριών ενός χρήστη η οποία εφαρμόστηκε σε πραγματικά ιατρικά δεδομένα. Η εξαγωγή των χαρακτηριστικών που διαφοροποιούνται μεταξύ των δύο καταστάσεων βασίστηκε σε πληροφορίες του πεδίου χρόνου-συχνότητας, οι οποίες αντλούνται με το φιλτράρισμα των ακατέργαστων ΗΕΓ δεδομένων και με τη βοήθεια των αιτιατών κυματιδίων Morlet, ενώ για την επακόλουθη ταξινόμηση των χαρακτηριστικών αναπτύχθηκαν και συγκρίθηκαν δύο αξιόπιστες μέθοδοι. Η πρώτη αφορά στη δημιουργία πιθανοθεωρητικών προτύπων κανονικής κατανομής για κάθε κατηγορία πρόθεσης κίνησης, με την τελική απόφαση ταξινόμησης να λαμβάνεται με εφαρμογή του απλού ταξινομητή του Bayes, ενώ η δεύτερη δημιουργεί ένα πρότυπο ταξινόμησης με βάση το θεωρητικό πλαίσιο των Μηχανών Διανυσμάτων Υποστήριξης (SVM). Στόχος του προβλήματος της δυαδικής ταξινόμησης είναι να αποφασίζεται σε ποια από τις δύο κατηγορίες ανήκει μία δεδομένη νοητική πρόθεση όσο το δυνατόν ταχύτερα και αξιόπιστα, έτσι ώστε ο σχεδιαζόμενος αλγόριθμος να εξυπηρετήσει ένα πλαίσιο ανατροφοδότησης της τελικής απόφασης στο χρήστη σε συνθήκες πραγματικού χρόνου. / Brain-Computer Interfaces (BCIs) demand the efficient processing of EEG data in order to translate one's thought or wish into a control signal that can be applied as input to external devices. Here we present a method to classify left from right hand movements, by extracting features from the data with Morlet wavelets and classifying with two different models, SVMs and Naive Bayes Classifier.

Thesis (MSc)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: Supervised classifiers are commonly used in remote sensing to extract land cover information. They are, however, limited in their ability to cost-effectively produce sufficiently accurate land cover maps. Various factors affect the accuracy of supervised classifiers. Notably, the number of available training samples is known to significantly influence classifier performance and to obtain a sufficient number of samples is not always practical. The support vector machine (SVM) does perform well with a limited number of training samples. But little research has been done to evaluate SVM’s performance for geographical object-based image analysis (GEOBIA). GEOBIA also allows the easy integration of additional features into the classification process, a factor which may significantly influence classification accuracies. As such, two experiments were developed and implemented in this research. The first compared the performances of object-based SVM, maximum likelihood (ML) and nearest neighbour (NN) classifiers using varying training set sizes. The effect of feature dimensionality on classifier accuracy was investigated in the second experiment. A SPOT 5 subscene and a four-class classification scheme were used. For the first experiment, training set sizes ranging from 4-20 per land cover class were tested. The performance of all the classifiers improved significantly as the training set size was increased. The ML classifier performed poorly when few (<10 per class) training samples were used and the NN classifier performed poorly compared to SVM throughout the experiment. SVM was the superior classifier for all training set sizes although ML achieved competitive results for sets of 12 or more training samples per class. Training sets were kept constant (20 and 10 samples per class) for the second experiment while an increasing number of features (1 to 22) were included. SVM consistently produced superior classification results. SVM and NN were not significantly (negatively) affected by an increase in feature dimensionality, but ML’s ability to perform under conditions of large feature dimensionalities and few training areas was limited. Further investigations using a variety of imagery types, classification schemes and additional features; finding optimal combinations of training set size and number of features; and determining the effect of specific features should prove valuable in developing more costeffective ways to process large volumes of satellite imagery. KEYWORDS Supervised classification, land cover, support vector machine, nearest neighbour classification maximum likelihood classification, geographic object-based image analysis / AFRIKAANSE OPSOMMING: Gerigte klassifiseerders word gereeld aangewend in afstandswaarneming om inligting oor landdekking te onttrek. Sulke klassifiseerders het egter beperkte vermoëns om akkurate landdekkingskaarte koste-effektief te produseer. Verskeie faktore het ʼn uitwerking op die akkuraatheid van gerigte klassifiseerders. Dit is veral bekend dat die getal beskikbare opleidingseenhede ʼn beduidende invloed op klassifiseerderakkuraatheid het en dit is nie altyd prakties om voldoende getalle te bekom nie. Die steunvektormasjien (SVM) werk goed met beperkte getalle opleidingseenhede. Min navorsing is egter gedoen om SVM se verrigting vir geografiese objek-gebaseerde beeldanalise (GEOBIA) te evalueer. GEOBIA vergemaklik die integrasie van addisionele kenmerke in die klassifikasie proses, ʼn faktor wat klassifikasie akkuraathede aansienlik kan beïnvloed. Twee eksperimente is gevolglik ontwikkel en geïmplementeer in hierdie navorsing. Die eerste eksperiment het objekgebaseerde SVM, maksimum waarskynlikheids- (ML) en naaste naburige (NN) klassifiseerders se verrigtings met verskillende groottes van opleidingstelle vergelyk. Die effek van kenmerkdimensionaliteit is in die tweede eksperiment ondersoek. ʼn SPOT 5 subbeeld en ʼn vier-klas klassifikasieskema is aangewend. Opleidingstelgroottes van 4-20 per landdekkingsklas is in die eerste eksperiment getoets. Die verrigting van die klassifiseerders het beduidend met ʼn toename in die grootte van die opleidingstelle verbeter. ML het swak presteer wanneer min (<10 per klas) opleidingseenhede gebruik is en NN het, in vergelyking met SVM, deurgaans swak presteer. SVM het die beste presteer vir alle groottes van opleidingstelle alhoewel ML kompeterend was vir stelle van 12 of meer opleidingseenhede per klas. Die grootte van die opleidingstelle is konstant gehou (20 en 10 eenhede per klas) in die tweede eksperiment waarin ʼn toenemende getal kenmerke (1 tot 22) toegevoeg is. SVM het deurgaans beter klassifikasieresultate gelewer. SVM en NN was nie beduidend (negatief) beïnvloed deur ʼn toename in kenmerkdimensionaliteit nie, maar ML se vermoë om te presteer onder toestande van groot kenmerkdimensionaliteite en min opleidingsareas was beperk. Verdere ondersoeke met ʼn verskeidenheid beelde, klassifikasie skemas en addisionele kenmerke; die vind van optimale kombinasies van opleidingstelgrootte en getal kenmerke; en die bepaling van die effek van spesifieke kenmerke sal waardevol wees in die ontwikkelling van meer koste effektiewe metodes om groot volumes satellietbeelde te prosesseer. TREFWOORDE Gerigte klassifikasie, landdekking, steunvektormasjien, naaste naburige klassifikasie, maksimum waarskynlikheidsklassifikasie, geografiese objekgebaseerde beeldanalise

abstract: Many products undergo several stages of testing ranging from tests on individual components to end-item tests. Additionally, these products may be further "tested" via customer or field use. The later failure of a delivered product may in some cases be due to circumstances that have no correlation with the product's inherent quality. However, at times, there may be cues in the upstream test data that, if detected, could serve to predict the likelihood of downstream failure or performance degradation induced by product use or environmental stresses. This study explores the use of downstream factory test data or product field reliability data to infer data mining or pattern recognition criteria onto manufacturing process or upstream test data by means of support vector machines (SVM) in order to provide reliability prediction models. In concert with a risk/benefit analysis, these models can be utilized to drive improvement of the product or, at least, via screening to improve the reliability of the product delivered to the customer. Such models can be used to aid in reliability risk assessment based on detectable correlations between the product test performance and the sources of supply, test stands, or other factors related to product manufacture. As an enhancement to the usefulness of the SVM or hyperplane classifier within this context, L-moments and the Western Electric Company (WECO) Rules are used to augment or replace the native process or test data used as inputs to the classifier. As part of this research, a generalizable binary classification methodology was developed that can be used to design and implement predictors of end-item field failure or downstream product performance based on upstream test data that may be composed of single-parameter, time-series, or multivariate real-valued data. Additionally, the methodology provides input parameter weighting factors that have proved useful in failure analysis and root cause investigations as indicators of which of several upstream product parameters have the greater influence on the downstream failure outcomes. / Dissertation/Thesis / Ph.D. Electrical Engineering 2011

Orientador: Aparecido Nilceu Marana / Banca: Roberto Marcondes Cesar Junior / Banca: Ivan Rizzo Guilherme / Resumo: O reconhecimento facial é uma das principais formas de identificação humana. Apesar das pesquisas em reconhecimento facial automático terem crescido substancialmente ao longo dos últimos 35 anos, identificar pessoas a partir da face continua sendo um desafio para as áreas de Visão Computacional e Reconhecimento de Padrões. Em função dos cenários variarem desde a identificação a partir de fotografias até o reconhecimento baseado em vídeos sem nenhum tipo de controle ao serem gravados, os maiores desafios estão relacionados à independência contra diferentes tipos de iluminação, pose e expressão. O objetivo desta dissertação é propor técnicas que possam contribuir para a melhoria dos sistemas de reconhecimento facial. A primeira técnica endereça o problema da iluminação através da fusão dos espectros visível e infravermelho da face. Através desta abordagem, as taxas de reconhecimento foram melhoradas em 2.07% enquanto a taxa de erro igual (EER) foi reduzida em 45.47%. A segunda técnica trata do caso da extração e classificação de características faciais. Ela propõe um novo modelo para reconhecimento facial através do uso de características extraídas por Histogramas Census e de uma técnica de reconhecimento de padrões baseada em Máquinas de Vetores de Suporte (SVMs). Este outro grupo de experimentos nos possibilitou aumentar a precisão do reconhecimento no teste FERET fa/fb em 0.5%. Além destes resultados, algumas contribuições adicionais deste trabalho que merecem ser destacadas são a análise da dependência estatística entre classificadores de espectros diferentes e considerações sobre o comportamento de uma única C-SVC SVM para identificação de pessoas de forma eficaz. / Abstract: Face recognition is one of the primary ways of human identification. Although researches on automated face recognition have broadly increased along the last 35 years, it remains a challenging task in the fields of Computer Vision and Pattern Recognition. As the scenarios varies from static and constrained photographs to uncontrolled video images, the challenging issues on automatic face recognition are usually related with variations in illumination, pose and expressions. The goal of this master thesis is to propose techniques for the improvement of face recognition systems. The first technique addresses the problem of illumination by fusing the visible and the infrared spectra of the face. With this approach the recognition rates were improved in 2.07% while the Equal Error Rate (EER) were reduced in 45.47%. The second technique addresses the issue of face features extraction and classification. It proposes a new framework for face recognition by using features extracted by Census Histograms and a pattern recognition technique based on Support Vector Machines (SVMs). This other group of experiments enabled us to increase the recognition accuracy in the FERET fa/fb test in 0.5%. Beyond these results, additional contributions of this work that deserve to be highlighted are the statistical dependency analysis between face recognition systems based on different spectra and a better comprehension about the behavior of a single C-SVC SVM to reliably predict faces identities. / Mestre

Submitted by DIEGO HENRIQUE EMYGDIO LÁZARO null (diegoemygdio@gmail.com) on 2016-06-27T15:28:11Z No. of bitstreams: 1 Aplicação de Máquinas de Vetores de Suporte na Identificação de Perfis de Alunos de acordo com Características da Teoria das Inteligências Múltiplas.pdf: 2758329 bytes, checksum: 02e2c2154153f7f78fdc32629f761d03 (MD5) / Rejected by Ana Paula Grisoto (grisotoana@reitoria.unesp.br), reason: Solicitamos que realize uma nova submissão seguindo a orientação abaixo: O arquivo submetido não contém o certificado de aprovação. A versão submetida por você é considerada a versão final da dissertação/tese, portanto não poderá ocorrer qualquer alteração em seu conteúdo após a aprovação. Corrija esta informação e realize uma nova submissão contendo o arquivo correto. Agradecemos a compreensão. on 2016-06-27T17:22:37Z (GMT) / Submitted by DIEGO HENRIQUE EMYGDIO LÁZARO null (diegoemygdio@gmail.com) on 2016-06-27T20:26:31Z No. of bitstreams: 1 Aplicação de Máquinas de Vetores de Suporte na Identificação de Perfis de Alunos de acordo com as Características das Inteligências Múltiplas.pdf: 2980004 bytes, checksum: d8b55bde9f111d6df2e3cc9a8db5e8e9 (MD5) / Rejected by Ana Paula Grisoto (grisotoana@reitoria.unesp.br), reason: Solicitamos que realize uma nova submissão seguindo a orientação abaixo: O arquivo submetido está sem a ficha catalográfica. A versão submetida por você é considerada a versão final da dissertação/tese, portanto não poderá ocorrer qualquer alteração em seu conteúdo após a aprovação. Corrija esta informação e realize uma nova submissão contendo o arquivo correto. Agradecemos a compreensão. on 2016-06-28T18:22:33Z (GMT) / Submitted by DIEGO HENRIQUE EMYGDIO LÁZARO null (diegoemygdio@gmail.com) on 2016-06-28T19:33:55Z No. of bitstreams: 1 Aplicação de Máquinas de Vetores de Suporte na Identificação de Perfis de Alunos de acordo com Características da Teoria das Inteligências Múltiplas.pdf: 2736602 bytes, checksum: 51b12df288fa6ceb2ba0e0a908303beb (MD5) / Approved for entry into archive by Ana Paula Grisoto (grisotoana@reitoria.unesp.br) on 2016-06-28T19:59:20Z (GMT) No. of bitstreams: 1 lazaro_dhe_me_sjrp.pdf: 2736602 bytes, checksum: 51b12df288fa6ceb2ba0e0a908303beb (MD5) / Made available in DSpace on 2016-06-28T19:59:20Z (GMT). No. of bitstreams: 1 lazaro_dhe_me_sjrp.pdf: 2736602 bytes, checksum: 51b12df288fa6ceb2ba0e0a908303beb (MD5) Previous issue date: 2016-05-31 / Nesta dissertação foi desenvolvido um mecanismo de classificação capaz de identificar o perfil de um aluno de acordo com características da teoria das inteligências múltiplas, baseado em Support Vector Machines (SVMs, sigla em inglês para Máquinas de Vetores de Suporte), métodos de agrupamento e balanceamento de classes. O objetivo dessa classificação consiste em permitir que os tutores responsáveis por gerar o material para aulas em ferramentas de apoio ao ensino à distância possam utilizar este método de classificação para direcionar o conteúdo ao aluno de forma a explorar sua inteligência múltipla predominante. Para realização dos experimentos, duas SVMs foram criadas, utilizando o método de classificação baseado em k problemas binários, que reduzem o problema de múltiplas classes a um conjunto de problemas binários. Os resultados obtidos durante as fases de treino e teste das SVMs foram apresentados em percentuais por meio de um algoritmo de agrupamento particionado. Esses percentuais ajudam a interpretar a classificação do perfil de acordo com as inteligências predominantes. Além disso, com o uso de métodos de balanceamento de classes, obteve-se melhora no desempenho do classificador, assim, aumentando a eficácia do mecanismo, pois, suas taxas de incorreções foram baixas. / In this work, it was developed a mechanism in order to classify students’ profiles according to the Theory of Multiple Intelligences, based on Support Vector Machines (SVMs), cluster methods and classes balancing. By using these classifications, tutors, who prepare materials for classes in specific tools for distance education purposes, are able to suggest contents for students so that they are able to explore their predominant multiple intelligence. To perform these experiments, SVMs were created by using classification methods based on binary problems that reduce multiple classes problems into a set of binary problems. The results generated during the training and the SVM test stages were presented in percentages by using partitioning clustering algorithm. These percentages are helpful for analysis of profiles classifications according to multiple intelligences. Besides that, by using classes balancing methods, it was possible to obtain improvements on the classifier performance and, consequently, the mechanism efficiency was increased as well, considering the fact that inaccuracy rates were low.

Made available in DSpace on 2016-12-12T20:12:32Z (GMT). No. of bitstreams: 1 PGEF15MA050.pdf: 5724520 bytes, checksum: 4aae5cece599a90d29686f6da99e729a (MD5) Previous issue date: 2015-07-31 / Over the last decades, advances in Earth Observation technology have been playing an important role for forest monitoring worldwide. A remarkable improvement in remotely sensed data is the refinement of the spatial resolution. The region-based classification and object-based image analysis (OBIA) appears to be the most appropriates approaches to extract information coming from high spatial resolution images. Such information is becoming even more frequently. However, the costs involved in acquiring proprieraly OBIA software licenses is often too high. Therefore, the use of open source softwares is envisaged. This study aimed to evaluate open source softwares in order to classify secondary successional forest stages of Ombrophilous Mixed Forest environments in Southern Brazil. Three test sites were selected in the mountainous region of Santa Catarina State (SC). We used scenes from the airborne system for acquisition and post-processing of images (SAAPI) with a spatial resolution of 0.39m. The dataset consists of orthorectified images containing three spectral bands in the visible range (i.e. 0.38 0.70&#956;m), three spectral bands in the near infrared (i.e. 0.76 0.78&#956;m) and a digital surface model. Three methodologies were developed using decision tree algorithms available at the following open source softwares: InterIMAGE, GeoDMA, WEKA and QGIS. We selected the support vector machine algorithm (SVM) available in Orfeo ToolBox Monteverdi software for both region-based and pixel-based classification. We also evaluate the SPT software in order to obtain the ideal set for segmentation parameters. Results show that the classification of secondary successional forest stages as well as other land use classes performed well. Kappa index ranged from 0.60 to 0.89. Conditional accuracy and both producer s and user s accuracy were higher than 0.5. The best overall accuracy results were found for initial forest stages while the worst performance was observed in the intermediate successional forest stage. Region based classification and OBIA performed better than pixel by pixel based classification. The generated classified maps reveal the applicability of this aproaches for extracting useful information from the SAAPI images. Such information can also provide useful information for forest resources monitoring practices at the state level / Nas últimas décadas, observou-se uma notável evolução das tecnologias espaciais destinadas ao monitoramento dos recursos florestais. Um significativo avanço nos dados de sensoriamento é o refinamento da resolução espacial. Com a crescente disponibilidade de imagens de alta resolução espacial, a classificação por regiões e a análise baseada em objeto (Object Based Image Analysis- OBIA) apresentam-se como abordagens mais adequadas para extrair informações dessas imagens. No entanto, os custos envolvidos na aquisição de licenças de aplicativos comerciais disponíveis para este propósito costuma ser demasiadamente alto. Assim, faz-se necessário avaliar o uso de aplicativos na modalidade open source. Este trabalho teve como objetivo avaliar aplicativos open source relacionadas à classificação baseada em regiões e à mineração de dados, para classificar estádios sucessionais de florestas secundárias da Floresta Ombrófila Mista (FOM) em três áreas-teste situadas na região serrana de Santa Catarina (SC). No processamento, utilizaram-se ortoimagens do Sistema Aerotransportado de Aquisição e Pós-processamento de Imagens (Airborne System for Acquisition and Post-processing of Images- SAAPI) com alta resolução espacial (0,39 m) obtidas no levantamento aerofotogramétrico de SC. Os dados consistem de três bandas no visível (0,38 - 0,70 &#956;m), três bandas no infravermelho próximo (0,76 - 0,78 &#956;m) e o Modelo Digital de Superfície. Três metodologias foram desenvolvidas utilizando mineração de dados com algoritmos de árvore de decisão, nos aplicativos InterIMAGE, GeoDMA, WEKA e QGIS. O algoritmo de máquina de vetor suporte (Support Vector Machine- SVM) foi selecionado para a classificação baseada em regiões e por pixel no software Orfeo ToolBox Monteverdi. Testou-se também o software SPT para avaliação e escolha automática dos parâmetros da segmentação das imagens. Os resultados se mostraram satisfatórios para classificar estádios sucessionais da FOM, assim como outras classes de uso e cobertura da terra. As classificações apresentaram um índice Kappa variando entre 0,6 e 0,89. A avaliação condicional das classes referentes aos estádios sucessionais (exatidão e Kappa condicional do produtor e usuário), no geral, foram superiores a 0,5, sendo os melhores resultados obtidos na identificação do estádio inicial e os piores para o estádio médio. A classificação baseada em regiões e a OBIA foram significantemente superiores à classificação pixel a pixel. Estes resultados demonstram o potencial dessas abordagens na extração de informações de imagens de alta resolução espacial, como os provenientes do recobrimento aéreo estadual, bem como, a possibilidade de fornecer subsídios para a implementação de políticas públicas e no monitoramento dos recursos florestais em nível estadual

Made available in DSpace on 2014-06-11T19:29:40Z (GMT). No. of bitstreams: 0 Previous issue date: 2009-06-19Bitstream added on 2014-06-13T18:07:45Z : No. of bitstreams: 1 chiachia_g_me_sjrp.pdf: 1197775 bytes, checksum: a782f5b01605aa2a8b8bb080a56b3cad (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / O reconhecimento facial é uma das principais formas de identificação humana. Apesar das pesquisas em reconhecimento facial automático terem crescido substancialmente ao longo dos últimos 35 anos, identificar pessoas a partir da face continua sendo um desafio para as áreas de Visão Computacional e Reconhecimento de Padrões. Em função dos cenários variarem desde a identificação a partir de fotografias até o reconhecimento baseado em vídeos sem nenhum tipo de controle ao serem gravados, os maiores desafios estão relacionados à independência contra diferentes tipos de iluminação, pose e expressão. O objetivo desta dissertação é propor técnicas que possam contribuir para a melhoria dos sistemas de reconhecimento facial. A primeira técnica endereça o problema da iluminação através da fusão dos espectros visível e infravermelho da face. Através desta abordagem, as taxas de reconhecimento foram melhoradas em 2.07% enquanto a taxa de erro igual (EER) foi reduzida em 45.47%. A segunda técnica trata do caso da extração e classificação de características faciais. Ela propõe um novo modelo para reconhecimento facial através do uso de características extraídas por Histogramas Census e de uma técnica de reconhecimento de padrões baseada em Máquinas de Vetores de Suporte (SVMs). Este outro grupo de experimentos nos possibilitou aumentar a precisão do reconhecimento no teste FERET fa/fb em 0.5%. Além destes resultados, algumas contribuições adicionais deste trabalho que merecem ser destacadas são a análise da dependência estatística entre classificadores de espectros diferentes e considerações sobre o comportamento de uma única C-SVC SVM para identificação de pessoas de forma eficaz. / Face recognition is one of the primary ways of human identification. Although researches on automated face recognition have broadly increased along the last 35 years, it remains a challenging task in the fields of Computer Vision and Pattern Recognition. As the scenarios varies from static and constrained photographs to uncontrolled video images, the challenging issues on automatic face recognition are usually related with variations in illumination, pose and expressions. The goal of this master thesis is to propose techniques for the improvement of face recognition systems. The first technique addresses the problem of illumination by fusing the visible and the infrared spectra of the face. With this approach the recognition rates were improved in 2.07% while the Equal Error Rate (EER) were reduced in 45.47%. The second technique addresses the issue of face features extraction and classification. It proposes a new framework for face recognition by using features extracted by Census Histograms and a pattern recognition technique based on Support Vector Machines (SVMs). This other group of experiments enabled us to increase the recognition accuracy in the FERET fa/fb test in 0.5%. Beyond these results, additional contributions of this work that deserve to be highlighted are the statistical dependency analysis between face recognition systems based on different spectra and a better comprehension about the behavior of a single C-SVC SVM to reliably predict faces identities.

Técnicas de Aprendizado de Máquina não-simbólicas, como Redes Neurais Artificiais, Máquinas de Vetores de Suporte e combinação de classificadores têm mostrado um bom desempenho quando utilizadas para análise de dados. A grande limitação dessas técnicas é a falta de compreensibilidade do conhecimento armazenado em suas estruturas internas. Esta Tese apresenta uma pesquisa realizada sobre métodos de extração de representações compreensíveis do conhecimento armazenado nas estruturas internas dessas técnicas não-simbólicas, aqui chamadas de caixa preta, durante seu processo de aprendizado. A principal contribuição desse trabalho é a proposta de um novo método pedagógico para extração de regras que expliquem o processo de classificação seguido por técnicas não-simbólicas. Esse novo método é baseado na otimização (maximização) da similaridade entre rankings de classificação produzidos por técnicas de Aprendizado de Máquina simbólicas e não simbólicas (de onde o conhecimento interno esta sendo extraído). Experimentos foram realizados com vários conjuntos de dados e os resultados obtidos sugerem um bom potencial para o método proposto / Non-symbolic Machine Learning techniques, like Artificial Neural Networks, Support Vector Machines and Ensembles of classifiers have shown a good performance when they are used in data analysis. The strong limitation regarding the use of these techniques is the lack of comprehensibility of the knowledge stored in their internal structure. This Thesis presents an investigation of methods capable of extracting comprehensible representations of the knowledge acquired by these non-symbolic techniques, here named black box, during their learning process. The main contribution of this work is the proposal of a new pedagogical method for rule extraction that explains the classification process followed by non-symbolic techniques. This new method is based on the optimization (maximization) of the similarity between classification rankings produced by symbolic and non-symbolic (from where the internal knowledge is being extracted) Machine Learning techniques. Experiments were performed for several datasets and the results obtained suggest a good potential of the proposed method