Orientador: Anderson de Rezende Rocha / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Computação / Made available in DSpace on 2018-08-23T17:47:21Z (GMT). No. of bitstreams: 1 Pires_Ramon_M.pdf: 4429324 bytes, checksum: 4e4015bc2131a1f1a5e8aa215f24d98f (MD5) Previous issue date: 2013 / Resumo: A Retinopatia Diabética (RD), complicação provocada pela diabetes, se manifesta por meio de diferentes lesões que possuem suas especificidades. Estas especificidades são exploradas na literatura como estratégia para representação, proporcionando uma discriminação satisfatória entre imagens de pacientes normais e doentes. No entanto, por estarem fortemente atrelada _as características visuais de cada anomalia, a detecção de lesões distintas exige abordagens distintas. Neste trabalho, apresentamos um arcabouço geral cujo objetivo é automatizar o procedimento de análise de imagens de fundo de olho. O trabalho é dividido em quatro etapas: avaliação de qualidade, detecção de lesões individuais, triagem e verificação de necessidade de consulta. Na primeira etapa, aplicamos diferentes técnicas de caracterização de imagens para avaliar a qualidade das imagens por meio de dois critérios: definição de campo e detecção de borramentos. Na segunda etapa deste trabalho, propomos a continuação de um trabalho anterior desenvolvido pelo nosso grupo, no qual foi aplicado um método unificado na tentativa de detecção de lesões distintas. No nosso método para detecção de qualquer lesão, exploramos diferentes alternativas de representação em baixo nível (extração densa e esparsa) e médio nível (técnicas de coding/pooling para sacolas de palavras visuais) objetivando o desenvolvimento de um conjunto eficaz de detectores de lesões individuais. As pontuações provenientes de cada detector de lesão, obtidas para cada imagem, representam uma descrição de alto nível, ponto fundamental para a terceira e a quarta etapas. Tendo em mãos um conjunto de dados descritos em alto nível (pontuações dos detectores individuais), propomos, na terceira etapa do trabalho, a aplicação de técnicas de fusão de dados para o desenvolvimento de um método de detecção de múltiplas lesões. A descrição em alto nível também é explorada na quarta etapa para o desenvolvimento de um método eficaz de avaliação de necessidade de encaminhamento a um oftalmologista no intervalo de um ano, visando evitar que o médico seja sobrecarregado, bem como dar prioridade a pacientes em estado urgente / Abstract: Diabetic Retinopathy (DR), a common complication caused by diabetes, manifests through deferent lesions that have their particularities. These particularities are explored in the literature as methods for representation, providing a satisfactory discrimination between healthy/diseased retinas. However, by being strongly linked to the visual characteristics of each anomaly, the detection of distinct lesions requires distinct approaches. In this work, we present a general framework whose objective is to automate the eye-fundus image analysis. The work comprises four steps: image quality assessment, DR-related lesion detection, screening, and referral. In the first step, we apply characterization techniques to assess image quality by two criteria: field definition and blur detection. In the second step of this work, we extend up a previous work of our group which explored a unified method for detecting distinct lesions in eye-fundus images. In our approach for detection of any lesion, we explore several alternatives for low-level (dense and sparse extraction) and mid-level (coding/pooling techniques of bag of visual words) representations, aiming at the development of an effective set of individual DR-related lesion detectors. The scores derived from each individual DR-related lesion, taken for each image, represent a high-level description, fundamental point for the third and fourth steps. Given a dataset described in high-level (scores from the individual detectors), we propose, in the third step of the work, the use of machine learning fusion techniques aiming at the development of a multi-lesion detection method. The high-level description is also explored in the fourth step for the development of an effective method for evaluating the necessity of referral of a patient to an ophthalmologist in the interval of one year, avoiding overloading medical specialist with simple cases as well as give priority to patients in an urgent state / Mestrado / Ciência da Computação / Mestre em Ciência da Computação

L'utilisation de commandes gestuelles est une nouvelle méthode d'interaction sur interface tactile. Une bonne méthode pour faciliter la mémorisation de ces commandes gestuelles est de laisser l'utilisateur les personnaliser. Ce contexte applicatif induit une situation d'apprentissage croisé, où l'utilisateur doit mémoriser le jeu de symboles elle système doit apprendre à reconnaître les différents symboles. Cela implique un certain nombre de contraintes, à la fois sur le système de reconnaissance de symboles ct sur le système de supervision de son apprentissage. Il faut par exemple que le classifieur puisse apprendre à partir de peu de données, continuer à apprendre pendant son utilisation et suivre toute évolution des données indéfiniment. Le superviseur doit quant à lui optimiser la coopération entre l'utilisateur et le système de reconnaissance pour minimiser les interactions tout en maximisant l'apprentissage. Cette thèse présente d'une part, le système d'apprentissage évolutif Evolve oo, capable d'apprendre rapidement il partir de peu de données et de suivre les changements de concepts. D'autre part, elle introduit le superviseur actif en-ligne lntuiSup qui permet d'optimiser la coopération entre le système et l'utilisateur, lors de l'utilisation de commandes gestuelles personnalisées notamment Evolve oo est un système d'inférence floue, capable d'apprendre rapidement grâce aux capacités génératrices des prémisses des règles, tout en permettant d'obtenir une précision élevée grâce aux capacités discriminantes des conclusions d'ordre un. L'intégration d'oubli dans le processus d'apprentissage permet de maintenir le gain de l'apprentissage indéfiniment, permettant ainsi l'ajout de classes à n'importe quel moment de l'utilisation du système ct garantissant son évolutivité « à vie». Le superviseur actif en-ligne lntuiSup permet d'optimiser les interactions avec l'utilisateur pour entraîner un système d'apprentissage lorsque l'utilisateur est dans la boucle. Il permet de faire évoluer la proportion de données que l'utilisateur doit étiqueter en fonction de la difficulté du problème et de l'évolution de l'environnement (changements de concepts). L'utilisation d'une méthode de« dopage» de l'apprentissage permet d'optimiser la répartition de ces interactions avec l'utilisateur pour maximiser leur impact sur l'apprentissage. / Using gesture commands is a new way of interacting with touch sensitive interfaces. In order to facilitate user memorization of several commands, it is essential to let the user customize the gestures. This applicative context gives rise to a crosslearning situation, where the user has to memorize the set of commands and the system has to learn and recognize the different gestures. This situation implies several requirements, from the recognizer and from the system that supervizes its learning process. For instance, the recognizer has to be able to learn from few data samples, to keep learning during its use and to follow indefinitely any change of the data now. The supervisor has to optimize the cooperation between the recognizer and the system to minimize user interactions while maximizing recognizer learning. This thesis presents on the one hand the evolving recognition system Evolve oo, that is capable of fast teaming from few data samples, and that follows concept drifts. On the other hand, this thesis also presents the on line active supervisor lntuiSup, that optimizes user-system cooperation when the user is in the training loop, as during customized gesture command use for instance. The evolving classifier Evolve oo is a fuzzy inference system that is fast learning thanks to the generative capacity of rule premises, and at the same time giving high precision thanks to the discriminative capacity of first order rule conclusion. The use of forgetting in the learning process allows to maintain the learning gain indefinitely, enabling class adding at any stage of system learning, and guaranteeing lifelong evolving capacity. The on line active supervisor IntuiSup optimizes user interactions to train a classifier when the user is in the training loop. The proportion of data that is labeled by the user evolves to adapt to problem difficulty and to follow environment evolution (concept drift s). The use of a boosting method optimizes the timing of user interactions to maximize their impact on classifier learning process.

We have a wide-range of classifiers today that are employed in numerous applications, from credit scoring to speech-processing, with great technical and commercial success. No classifier, however, exists that will outperform all other classifiers on all classification tasks, and the process of classifier selection is still mainly one of trial and error. The optimal classifier for a classification task is determined by the characteristics of the data set employed; understanding the relationship between data characteristics and the performance of classifiers is therefore crucial to the process of classifier selection. Empirical and theoretical approaches have been employed in the literature to define this relationship. None of these approaches have, however, been very successful in accurately predicting or explaining classifier performance on real-world data. We use theoretical properties of classifiers to identify data characteristics that influence classifier performance; these data properties guide us in the development of measures that describe the relationship between data characteristics and classifier performance. We employ these data measures on real-world and artificial data to construct a meta-classification system. We use theoretical properties of classifiers to identify data characteristics that influence classifier performance; these data properties guide us in the development of measures that describe the relationship between data characteristics and classifier performance. We employ these data measures on real-world and artificial data to construct a meta-classification system. The purpose of this meta-classifier is two-fold: (1) to predict the classification performance of real-world classification tasks, and (2) to explain these predictions in order to gain insight into the properties of real-world data. We show that these data measures can be employed successfully to predict the classification performance of real-world data sets; these predictions are accurate in some instances but there is still unpredictable behaviour in other instances. We illustrate that these data measures can give valuable insight into the properties and data structures of real-world data; these insights are extremely valuable for high-dimensional classification problems. / Dissertation (MEng)--University of Pretoria, 2008. / Electrical, Electronic and Computer Engineering / unrestricted

Afin de traiter les effluents liquides radioactifs de faible et moyenne activités, le procédé utilisé à l'échelle industrielle est la coprécipitation. L'enjeu de cette thèse est d'optimiser le procédé continu de coprécipitation. Pour cela, un réacteur innovant est conçu et modélisé : le réacteur/décanteur continu. Deux systèmes modèles sont étudiés : la coprécipitation du strontium par le sulfate de baryum et la sorption du césium par le PPFeNi. Le milieu étudié est une solution contenant du nitrate de sodium afin de prendre en compte la force ionique élevée des effluents. Chaque système modèle est d'abord étudié de manière séparée, puis de manière simultanée. Les lois cinétiques de nucléation et de croissance cristalline du sulfate de baryum sont déterminées, puis intégrées au modèle de coprécipitation. Des études de cinétique et d'isotherme de sorption du césium par le PPFeNi sont aussi menées afin d'acquérir les données nécessaires à la modélisation du procédé. La modélisation permet de prédire finement la concentration résiduelle en strontium et en césium en fonction du type de procédé utilisé : cela constitue un outil précieux pour l'optimisation d'unités existantes ou le dimensionnement d'unités futures. Le réacteur/décanteur continu présente de très nombreux avantages par rapport au procédé continu classique : il permet d'améliorer sensiblement les performances de décontamination en strontium et en césium tout en réduisant le volume de boues générées par le procédé. Le réacteur/décanteur assure aussi une bonne séparation liquide/solide, et l'installation résultante se révèle nettement plus compacte. Ainsi, le réacteur/décanteur continu permet d'intensifier les procédés de traitement d'effluents liquides radioactifs, et constitue une technologie très prometteuse pour une application industrielle future / In order to decontaminate radioactive liquid wastes of low and intermediate levels, the coprecipitation is the process industrially used. The aim of this PhD work is to optimize the continuous process of coprecipitation. To do so, an innovative reactor is designed and modelled: the continuous reactor/classifier. Two model systems are studied: the coprecipitation of strontium by barium sulphate and the sorption of cesium by PPFeNi. The simulated effluent contains sodium nitrate in order to consider the high ionic strength of radioactive liquid wastes. First, each model system is studied on its own, and then a simultaneous treatment is performed. The kinetic laws of nucleation and crystal growth of barium sulphate are determined and incorporated into the coprecipitation model. Kinetic studies and sorption isotherms of cesium by PPFeNi are also performed in order to acquire the necessary data for process modelling. The modelling realised enables accurate prediction of the residual strontium and cesium concentrations according to the process used: it is a valuable tool for the optimization of existing units, but also the design of future units. The continuous reactor/classifier presents many advantages compared to the classical continuous process: the decontamination efficiency of strontium and cesium is highly improved while the volume of sludge generated by the process is reduced. A better liquid/solid separation is observed in the reactor/classifier and the global installation is significantly more compact. Thus, the radioactive liquid wastes treatment processes can be intensified by the continuous reactor/classifier, which represents a very promising technology for future industrial application

[pt] Esta dissertação investigou a criação de comitês de classificadores baseados em Redes Neurais Multilayer Perceptron (Redes MLP, abreviadamente). Isso foi feito em dois passos: primeiro, aplicando-se procedimentos para criação de redes complementares, i.e, redes individualmente eficazes mas que cometem erros distintos; segundo, testando- se sobre essas redes alguns dos principais métodos de combinação disponíveis. Dentre os procedimentos para criação de redes complementares, foi dado enfoque para os baseados em alteração do conjunto de treinamento. Os métodos Bootstrap e Arc-x4 foram escolhidos para serem utilizados no estudo de casos, juntamente com o método RDP (Replicação Dirigida de Padrões). No que diz respeito aos métodos de combinação disponíveis, foi dada particular atenção ao método de combinação por integrais nebulosas. Além deste método, implementou-se combinação por média, votação por pluralidade e Borda count. As aplicações escolhidas para teste envolveram duas vertentes importantes na área de visão computacional - Classificação de Coberturas de Solo por Imagens de Satélite e Reconhecimento de Expressões Faciais. Embora ambas pertençam à mesma área de conhecimento, foram escolhidas de modo a representar níveis de dificuldade diferentes como tarefas de classificação - enquanto a primeira contou com um grande número de padrões disponíveis, a segunda foi comparativamente limitada nesse sentido. Como resultado final, comprovou-se a viabilidade da utilização de comitês em problemas de classificação, mesmo com as possíveis variações de desempenho relacionadas com a complexidade desses problemas. O método de combinação baseado em integrais nebulosas mostrou-se particularmente eficiente quando associado ao procedimento RDP para formação das redes comissionadas, mas nem sempre foi satisfatório. Considerado individualmente, o RDP tem a limitação de criar, no máximo, tantas redes quanto forem as classes consideradas em um problema; porém, quando este número de redes foi considerado como base de comparação, o RDP se mostrou, na média de todos os métodos de combinação testados, mais eficaz que os procedimentos Bootstrap e Arc-x4. Por outro lado, tanto o Bootstrap quanto o Arc-x4 têm a importante vantagem de permitirem a formação de um número crescente de membros, o que quase sempre acarretou em melhorias de desempenho global em relação ao RDP. / [en] The present dissertation investigated the creation of classifier committees based on Multilayer Perceptron Neural Networks (MLP Networks, for short). This was done in two parts: first, by applying procedures for creating complementary networks, i.e., networks that are individually accurate but cause distinct misclassifications; second, by assessing different combining methods to these network`s outputs. Among the procedures for creating committees members, the main focus was set to the ones based on changes to the training set . Bootstrap and Arc-x4 were chosen to be used at the experiments, along with the RDP procedure (translated as Driven Pattern Replication). With respect to the available combining methods, special attention was paid to fuzzy integrals combination. Average combination, plurality voting and Borda count were also implemented. The chosen experimental applications included interesting branches from computer vision: Land Cover Classification from Satellite Images and Facial Expression Recognition. These applications were specially interesting, in the sense they represent two different levels of difficulty as classification tasks - while the first had a great number of available patterns, the second was comparatively limited in this way. This work proved the viability of using committees in classification problems, despite the small performance fluctuations related to these problems complexity. The fuzzy integrals method has shown to be particularly interesting when coupled with the RDP procedure for committee creation, but was not always satisfactory. Taken alone, the RDP has the limitation of creating, at most, as many networks as there are classes to be considered at the problem at hand; however, when this number of networks was considered as the basis for comparison, this procedure outperformed, taking into account average combining results, both Bootstrap and Arc- x4. On the other hand, these later procedures have the important advantage of allowing the creation of an increasing number of committee members, what almost always increased global performance in comparison to RDP. / [es] Esta disertación investigó la creación de comités de clasificadores basados en Redes Neurales Multilayer Perceptron (Redes MLP, abreviadamente). Esto fue ejecutado en dos pasos: primeiro, aplicando procedimentos para la creación de redes complementares, esto es, redes que individualmente son eficaces pero que cometen erros diferentes; segundo, probando sobre esas redes, algunos de los principales métodos de combinación disponibles. Dentro de los procedimentos para la creación de redes complementares, se eligieron los basados en alteración del conjunto de entrenamiento. Los métodos Bootstrap y Arc-x4 fueron seleccionados para utilizarlos em el estudio de casos, conjuntamente con el método RDP (Replicación Dirigida de Padrones). Con respecto a los métodos de combinación disponibles, se le dió particular atención al método de combinación por integrales nebulosas. Además de este método, se implementaron: combinación por media, votación por pluralidad y Borda cont. Las aplicaciones seleccionadas para pruebas consideran dos vertientes importantes en la área de visión computacional - Clasificación de Coberturas de Suelo por Imágenes de Shastalite y Reconocimiento de Expresiones Faciales. Aunque ambas pertencen a la misma área de conocimento, fueron seleccionadas de modo con diferentes níveles de dificuldad como tareas de clasificación - Mientras la primera contó con un gran número de padrones disponibles, la segunda fue comparativamente limitada em ese sentido. Como resultado final, se comprobó la viabilidad de la utilización de comités en problemas de clasificación, incluso con las posibles variaciones de desempeño relacionadas con la complejidad de esos problemas. El método de combinación basado en integrales nebulosas se mostró particularmente eficiente asociado al procedimiento RDP para formación de las redes comisionadas, pero no siempre fue satisfactorio. Considerado individualmente, el RDP tiene la limitación de crear, como máximo, tantas redes como clases consideradas en un problema; sin embargo, cuando el número de redes fue considerado como base de comparación, el RDP se mostró más eficaz, en la media de todos los métodos de combinación, que los procedimentos Bootstrap y Arc-x4. Por otro lado, tanto el Bootstrap como el Arc-x4 tiene la importante ventaja de permitir la formación de un número cresciente de miembros, lo que generalmente mejora el desempeño global en relación al RDP.

No description available.

Tato diplomová práce pojednává o dvou algoritmech pro dolování z proudu dat - Very Fast Decision Tree (VFDT) a Concept-adapting Very Fast Decision Tree (CVFDT). Je vysvětlen princip klasifikace rozhodovacím stromem. Je popsána základní myšlenka konstrukce stromu Hoeffding Tree, který je základem pro algoritmy VFDT a CVFDT. Tyto algoritmy jsou poté rozebrány detailněji. Dále se tato práce zabývá návrhem algoritmu Genetického Programování (GP), který je použit pro vytváření klasifikátoru obrazových dat. Vytvořený klasifikátor je použit jako alternativní způsob klasifikace objektů v obraze ve frameworku Viola-Jones. V práci je rozebrána implementace algoritmů, které jsou implementovány v jazyce Java. Algoritmus GP je integrován do knihovny “Image Processing Extension” programu RapidMiner. Algoritmy VFDT a CVFDT jsou testovány na syntetických a reálných textových datech. Algoritmus GP je testován na klasifikaci obrazových dat a následně vytvořený klasifikátor je otestován na detekci obličejů v obraze.

This work deals with creating a hybrid system based on the aggregation of artificial immune system with appropriate heuristics to make the most effective spam detection. This work describes the main principles of biological and artificial immune system and conventional techniques to detect spam including several classifiers. The developed system is tested using well known database corpuses and a comparison of the final experiments is made.

In today’s modern digital world more and more emails and messengers must be sent, processed and handled. The categorizing and classification of these text pieces can take an incredibly long time and will cost the company a lot of time and money. If the classification could be done automatically by a computer dependent on the content of the text/message it would result in a major yield for the Easit AB and its customers. In order to facilitate the task of text-classification Easit needs a solution that is made out of one language model and one classifier model. The language model will convert raw text to a vector that is representative of the text and the classifier will construe what predefined labels fit for the vector. The end goal is not to create the best solution. It is simply to create a general understanding about different language and classifier models and how to build a system that will be both fast and accurate. BERT were the primary language model during evaluation but doc2Vec and One-Hot encoding was also tested. The classifier consisted out of boundary condition models or dense neural networks that were all trained without knowledge about what language model that the text vectors came from. The validation accuracy which was presented for the IMDB-comment dataset with BERT resulted between 75% to 94%, mostly dependent on the language model and not on the classifier. The knowledge from the work resulted in a recommendation to Easit for an alternativebased system solution. / I dagens moderna digitala värld är det allt mer majl-ärenden och meddelanden som ska skickas och processeras. Kategorisering och klassificering av dessa kan ta otroligt lång tid och kostar företag tid samt pengar. Om klassifieringen kunde ske automatiskt beroende på text-innehållet skulle det innebära en stor vinst för Easit AB och deras kunder.  För att underlätta arbetet med text-klassifiering behöver Easit en tvådelad lösning som består utav en språkmodell och en klassifierare. Språkmodellen som omvandlar text till en vektor som representerar texten och klassifieraren tolkar vilka fördefinerade ettiketter/märken som passar för vektorn. Målet är inte att skapa den bästa lösningen utan det är att skapa en generell kunskap för hur man kan utforma ett system som kan klassifiera texten på ett träffsäkert och effektivt sätt. Vid utvärdering av olika språkmodeller användes framförallt BERT-modeller men även doc2Vec och One-Hot testas också. Klassifieraren bestod utav gränsvillkors-modeller eller dense neurala nätverk som tränades helt utan vetskap om vilken språkmodell som skickat text-vektorerna. Träffsäkerheten som uppvisades vid validering för IMDB-kommentars datasetet med BERT blev mellan 75% till 94%, primärt beroende på språkmodellen. De neuralt nätverk passar bäst som klassifierare mest på grund av deras skalbarhet med flera ettiketter. Kunskapen från arbetet resulterade i en rekommendation till Easit om en alternativbaserad systemlösning.

Research philosophies and paradigms (RPPs) reveal researchers’ assumptions and provide a systematic way in which research can be carried out effectively and appropriately. Different studies highlight cognitive and comprehension challenges of RPPs concepts at the postgraduate level. This study develops a natural language processing (NLP) supervised classification application that guides students in identifying RPPs applicable to their study. By using algorithms rooted in a quantitative research approach, this study builds a corpus represented using the Bag of Words model to train the naïve Bayes, Logistic Regression, and Support Vector Machine algorithms. Computer experiments conducted to evaluate the performance of the algorithms reveal that the Naïve Bayes algorithm presents the highest accuracy and precision levels. In practice, user testing results show the varying impact of knowledge, performance, and effort expectancy. The findings contribute to the minimization of issues postgraduates encounter in identifying research philosophies and the underlying paradigms for their studies. / Science and Technology Education / MTech. (Information Technology)