Made available in DSpace on 2014-06-12T16:01:14Z (GMT). No. of bitstreams: 2 arquivo7274_1.pdf: 1733810 bytes, checksum: 2d9eb7a4489382e5afbf1790810474a0 (MD5) license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2005 / A análise de dados simbólicos (Symbolic Data Analysis) é um novo domínio na área de descoberta automática de conhecimento que visa desenvolver métodos para dados descritos por variáveis que podem assumir como valor conjuntos de categorias, intervalos ou distribuições de probabilidade. Essas novas variáveis permitem levar em conta a variabilidade e/ou a incerteza presente nos dados. O tratamento de dados simbólicos através de técnicas estatísticas e de aprendizagem de máquina necessita da introdução de medidas de distância capazes de manipular tal tipo de dado. Com esse objetivo, diversas funções de dissimilaridade têm sido propostas na literatura. Entretanto, nenhum estudo comparativo acerca do desempenho de tais funções em problemas que envolvem simultaneamente dados simbólicos booleanos e modais foi realizado. A principal contribuição dessa dissertação é realizar uma análise comparativa e uma avaliação empírica sobre funções de dissimilaridade para dados simbólicos, uma vez que esse tipo de estudo, apesar de muito relevante, é quase inexistente na literatura. Além disso, este trabalho também introduz novas funções de dissimilaridade que podem ser usadas no agrupamento dinâmico de dados simbólicos. Os algoritmos de agrupamento dinâmico consistem em obter, simultaneamente, uma partição em um número fixo de classes e a identificação de um representante para cada classe, minimizando localmente um critério que mede a adequação entre as classes e os seus representantes. Para validar esse estudo, foram realizados experimentos com bases de dados de referência na literatura e dois conjuntos de dados artificiais de intervalos com diferentes graus de dificuldade de classificação, objetivando a comparação das funções avaliadas. A precisão dos resultados foi mensurada por um índice externo de agrupamento aplicado na validação cruzada não supervisionada, para as bases de dados reais, e também no quadro de uma experiência Monte Carlo, para as bases de dados artificiais. Com os resultados alcançados é possível verificar a adequação das diversas funções de dissimilaridade aos diferentes tipos de dados simbólicos (multivalorado, multivalorado ordinal, intervalar, e modal de mesmo suporte e de suportes diferentes), bem como identificar as melhores configurações de funções. Testes estatísticos validam as conclusões

L'objectif principal de cette thèse est de développer un algorithme dynamique de partitionnement de données (classification non supervisée ou " clustering " en anglais) qui ne se limite pas à des concepts statiques et qui peut gérer des distributions qui évoluent au fil du temps. Cet algorithme peut être utilisé dans les systèmes de surveillance du processus, mais son application ne se limite pas à ceux-ci. Les contributions de cette thèse peuvent être présentées en trois groupes: 1. Contributions au partitionnement dynamique de données en utilisant : un algorithme de partitionnement dynamique basé à la fois sur la distance et la densité des échantillons est présenté. Cet algorithme ne fait aucune hypothèse sur la linéarité ni la convexité des groupes qu'il analyse. Ces clusters, qui peuvent avoir des densités différentes, peuvent également se chevaucher. L'algorithme développé fonctionne en ligne et fusionne les étapes d'apprentissage et de reconnaissance, ce qui permet de détecter et de caractériser de nouveaux comportements en continu tout en reconnaissant l'état courant du système. 2. Contributions à l'extraction de caractéristiques : une nouvelle approche permettant d'extraire des caractéristiques dynamiques est présentée. Cette approche, basée sur une approximation polynomiale par morceaux, permet de représenter des comportements dynamiques sans perdre les informations relatives à la magnitude et en réduisant simultanément la sensibilité de l'algorithme au bruit dans les signaux analysés. 3. Contributions à la modélisation de systèmes à événements discrets évolutifs a partir des résultats du clustering : les résultats de l'algorithme de partitionnement sont utilisés comme base pour l'élaboration d'un modèle à événements discrets du processus. Ce modèle adaptatif offre une représentation du comportement du processus de haut niveau sous la forme d'un automate dont les états représentent les états du processus appris par le partitionnement jusqu'à l'instant courant et les transitions expriment l'atteignabilité des états. / The main objective of this thesis is to propose a dynamic clustering algorithm that can handle not only dynamic data but also evolving distributions. This algorithm is particularly fitted for the monitoring of processes generating massive data streams, but its application is not limited to this domain. The main contributions of this thesis are: 1. Contribution to dynamic clustering by the proposal of an approach that uses distance- and density-based analyses to cluster non-linear, non-convex, overlapped data distributions with varied densities. This algorithm, that works in an online fashion, fusions the learning and lassification stages allowing to continuously detect and characterize new concepts and at the same time classifying the input samples, i.e. which means recognizing the current state of the system in a supervision application. 2. Contribution to feature extraction by the proposal of a novel approach to extract dynamic features. This approach ,based on piece-polynomial approximation, allows to represent dynamic behaviors without losing magnitude related information and to reduce at the same time the algorithm sensitivity to noise corrupting the signals. 3. Contribution to automatic discrete event modeling for evolving systems by exploiting informations brought by the clustering. The generated model is presented as a timed automaton that provides a high-level representation of the behavior of the process. The latter is adaptive in the sense that its construction is elaborated following the discovery of new concepts by the clustering algorithm.

京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第20032号 / 情博第627号 / 新制||情||109(附属図書館) / 33128 / 京都大学大学院情報学研究科通信情報システム専攻 / (主査)教授 原田 博司, 教授 守倉 正博, 教授 梅野 健 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

No description available.

[EN] The electricity sector is currently undergoing a process of liberalization and separation of roles, which is being implemented under the regulatory auspices of each Member State of the European Union and, therefore, with different speeds, perspectives and objectives that must converge on a common horizon, where Europe will benefit from an interconnected energy market in which producers and consumers can participate in free competition. This process of liberalization and separation of roles involves two consequences or, viewed another way, entails a major consequence from which other immediate consequence, as a necessity, is derived. The main consequence is the increased complexity in the management and supervision of a system, the electrical, increasingly interconnected and participatory, with connection of distributed energy sources, much of them from renewable sources, at different voltage levels and with different generation capacity at any point in the network. From this situation the other consequence is derived, which is the need to communicate information between agents, reliably, safely and quickly, and that this information is analyzed in the most effective way possible, to form part of the processes of decision taking that improve the observability and controllability of a system which is increasing in complexity and number of agents involved. With the evolution of Information and Communication Technologies (ICT), and the investments both in improving existing measurement and communications infrastructure, and taking the measurement and actuation capacity to a greater number of points in medium and low voltage networks, the availability of data that informs of the state of the network is increasingly higher and more complete. All these systems are part of the so-called Smart Grids, or intelligent networks of the future, a future which is not so far. One such source of information comes from the energy consumption of customers, measured on a regular basis (every hour, half hour or quarter-hour) and sent to the Distribution System Operators from the Smart Meters making use of Advanced Metering Infrastructure (AMI). This way, there is an increasingly amount of information on the energy consumption of customers, being stored in Big Data systems. This growing source of information demands specialized techniques which can take benefit from it, extracting a useful and summarized knowledge from it. This thesis deals with the use of this information of energy consumption from Smart Meters, in particular on the application of data mining techniques to obtain temporal patterns that characterize the users of electrical energy, grouping them according to these patterns in a small number of groups or clusters, that allow evaluating how users consume energy, both during the day and during a sequence of days, allowing to assess trends and predict future scenarios. For this, the current techniques are studied and, proving that the current works do not cover this objective, clustering or dynamic segmentation techniques applied to load profiles of electric energy consumption from domestic users are developed. These techniques are tested and validated on a database of hourly energy consumption values for a sample of residential customers in Spain during years 2008 and 2009. The results allow to observe both the characterization in consumption patterns of the different types of residential energy consumers, and their evolution over time, and to assess, for example, how the regulatory changes that occurred in Spain in the electricity sector during those years influenced in the temporal patterns of energy consumption. / [ES] El sector eléctrico se halla actualmente sometido a un proceso de liberalización y separación de roles, que está siendo aplicado bajo los auspicios regulatorios de cada Estado Miembro de la Unión Europea y, por tanto, con distintas velocidades, perspectivas y objetivos que deben confluir en un horizonte común, en donde Europa se beneficiará de un mercado energético interconectado, en el cual productores y consumidores podrán participar en libre competencia. Este proceso de liberalización y separación de roles conlleva dos consecuencias o, visto de otra manera, conlleva una consecuencia principal de la cual se deriva, como necesidad, otra consecuencia inmediata. La consecuencia principal es el aumento de la complejidad en la gestión y supervisión de un sistema, el eléctrico, cada vez más interconectado y participativo, con conexión de fuentes distribuidas de energía, muchas de ellas de origen renovable, a distintos niveles de tensión y con distinta capacidad de generación, en cualquier punto de la red. De esta situación se deriva la otra consecuencia, que es la necesidad de comunicar información entre los distintos agentes, de forma fiable, segura y rápida, y que esta información sea analizada de la forma más eficaz posible, para que forme parte de los procesos de toma de decisiones que mejoran la observabilidad y controlabilidad de un sistema cada vez más complejo y con más agentes involucrados. Con el avance de las Tecnologías de Información y Comunicaciones (TIC), y las inversiones tanto en mejora de la infraestructura existente de medida y comunicaciones, como en llevar la obtención de medidas y la capacidad de actuación a un mayor número de puntos en redes de media y baja tensión, la disponibilidad de datos sobre el estado de la red es cada vez mayor y más completa. Todos estos sistemas forman parte de las llamadas Smart Grids, o redes inteligentes del futuro, un futuro ya no tan lejano. Una de estas fuentes de información proviene de los consumos energéticos de los clientes, medidos de forma periódica (cada hora, media hora o cuarto de hora) y enviados hacia las Distribuidoras desde los contadores inteligentes o Smart Meters, mediante infraestructura avanzada de medida o Advanced Metering Infrastructure (AMI). De esta forma, cada vez se tiene una mayor cantidad de información sobre los consumos energéticos de los clientes, almacenada en sistemas de Big Data. Esta cada vez mayor fuente de información demanda técnicas especializadas que sepan aprovecharla, extrayendo un conocimiento útil y resumido de la misma. La presente Tesis doctoral versa sobre el uso de esta información de consumos energéticos de los contadores inteligentes, en concreto sobre la aplicación de técnicas de minería de datos (data mining) para obtener patrones temporales que caractericen a los usuarios de energía eléctrica, agrupándolos según estos mismos patrones en un número reducido de grupos o clusters, que permiten evaluar la forma en que los usuarios consumen la energía, tanto a lo largo del día como durante una secuencia de días, permitiendo evaluar tendencias y predecir escenarios futuros. Para ello se estudian las técnicas actuales y, comprobando que los trabajos actuales no cubren este objetivo, se desarrollan técnicas de clustering o segmentación dinámica aplicadas a curvas de carga de consumo eléctrico diario de clientes domésticos. Estas técnicas se prueban y validan sobre una base de datos de consumos energéticos horarios de una muestra de clientes residenciales en España durante los años 2008 y 2009. Los resultados permiten observar tanto la caracterización en consumos de los distintos tipos de consumidores energéticos residenciales, como su evolución en el tiempo, y permiten evaluar, por ejemplo, cómo influenciaron en los patrones temporales de consumos los cambios regulatorios que se produjeron en España en el sector eléctrico durante esos años. / [CA] El sector elèctric es troba actualment sotmès a un procés de liberalització i separació de rols, que s'està aplicant davall els auspicis reguladors de cada estat membre de la Unió Europea i, per tant, amb distintes velocitats, perspectives i objectius que han de confluir en un horitzó comú, on Europa es beneficiarà d'un mercat energètic interconnectat, en el qual productors i consumidors podran participar en lliure competència. Aquest procés de liberalització i separació de rols comporta dues conseqüències o, vist d'una altra manera, comporta una conseqüència principal de la qual es deriva, com a necessitat, una altra conseqüència immediata. La conseqüència principal és l'augment de la complexitat en la gestió i supervisió d'un sistema, l'elèctric, cada vegada més interconnectat i participatiu, amb connexió de fonts distribuïdes d'energia, moltes d'aquestes d'origen renovable, a distints nivells de tensió i amb distinta capacitat de generació, en qualsevol punt de la xarxa. D'aquesta situació es deriva l'altra conseqüència, que és la necessitat de comunicar informació entre els distints agents, de forma fiable, segura i ràpida, i que aquesta informació siga analitzada de la manera més eficaç possible, perquè forme part dels processos de presa de decisions que milloren l'observabilitat i controlabilitat d'un sistema cada vegada més complex i amb més agents involucrats. Amb l'avanç de les tecnologies de la informació i les comunicacions (TIC), i les inversions, tant en la millora de la infraestructura existent de mesura i comunicacions, com en el trasllat de l'obtenció de mesures i capacitat d'actuació a un nombre més gran de punts en xarxes de mitjana i baixa tensió, la disponibilitat de dades sobre l'estat de la xarxa és cada vegada major i més completa. Tots aquests sistemes formen part de les denominades Smart Grids o xarxes intel·ligents del futur, un futur ja no tan llunyà. Una d'aquestes fonts d'informació prové dels consums energètics dels clients, mesurats de forma periòdica (cada hora, mitja hora o quart d'hora) i enviats cap a les distribuïdores des dels comptadors intel·ligents o Smart Meters, per mitjà d'infraestructura avançada de mesura o Advanced Metering Infrastructure (AMI). D'aquesta manera, cada vegada es té una major quantitat d'informació sobre els consums energètics dels clients, emmagatzemada en sistemes de Big Data. Aquesta cada vegada major font d'informació demanda tècniques especialitzades que sàpiguen aprofitar-la, extraient-ne un coneixement útil i resumit. La present tesi doctoral versa sobre l'ús d'aquesta informació de consums energètics dels comptadors intel·ligents, en concret sobre l'aplicació de tècniques de mineria de dades (data mining) per a obtenir patrons temporals que caracteritzen els usuaris d'energia elèctrica, agrupant-los segons aquests mateixos patrons en una quantitat reduïda de grups o clusters, que permeten avaluar la forma en què els usuaris consumeixen l'energia, tant al llarg del dia com durant una seqüència de dies, i que permetent avaluar tendències i predir escenaris futurs. Amb aquesta finalitat, s'estudien les tècniques actuals i, en comprovar que els treballs actuals no cobreixen aquest objectiu, es desenvolupen tècniques de clustering o segmentació dinàmica aplicades a corbes de càrrega de consum elèctric diari de clients domèstics. Aquestes tècniques es proven i validen sobre una base de dades de consums energètics horaris d'una mostra de clients residencials a Espanya durant els anys 2008 i 2009. Els resultats permeten observar tant la caracterització en consums dels distints tipus de consumidors energètics residencials, com la seua evolució en el temps, i permeten avaluar, per exemple, com van influenciar en els patrons temporals de consums els canvis reguladors que es van produir a Espanya en el sector elèctric durant aquests anys. / Benítez Sánchez, IJ. (2015). Dynamic segmentation techniques applied to load profiles of electric energy consumption from domestic users [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/59236

Submitted by Fabio Sobreira Campos da Costa (fabio.sobreira@ufpe.br) on 2016-08-08T12:52:58Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) teseCinLeandro.pdf: 1316077 bytes, checksum: 61e762c7526a38a80ecab8f5c7769a47 (MD5) / Made available in DSpace on 2016-08-08T12:52:58Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) teseCinLeandro.pdf: 1316077 bytes, checksum: 61e762c7526a38a80ecab8f5c7769a47 (MD5) Previous issue date: 2016-01-18 / Um intervalo é um tipo de dado complexo usado na agregação de informações ou na representação de dados imprecisos. Este trabalho apresenta duas novas representações para intervalos com o objetivo de se construir novos métodos de agrupamento e regressão linear para este tipo de dado. O agrupamento por nuvens dinâmicas define partições nos dados e associa protótipos a cada uma destas partições. Os protótipos resumem a informação das partições e são usados na minimização de um critério que depende de uma distância, responsável por quantificar a proximidade entre instâncias e protótipos. Neste sentido, propõe-se a formulação de uma nova distância híbrida entre intervalos baseando-se em distâncias para pontos. Os pontos utilizados são obtidos dos intervalos através de um mapeamento. Também são propostas duas versões com pesos para a distância criada: uma com pesos no hibridismo e outra com pesos adaptativos. Na regressão linear, propõe-se a representação dos intervalos através da equação paramétrica da reta. Esta parametrização permite o ajuste dos pontos nas variáveis regressoras que dão as melhores estimativas para os limites da variável resposta. Antes da realização da regressão, um critério é calculado para a verificação da coerência matemática da predição, na qual o limite superior deve ser maior ou igual ao inferior. Se o critério mostra que a coerência não é garantida, propõe-se a aplicação de uma transformação sobre a variável resposta. Assim, este trabalho também propõe algumas transformações que podem ser aplicadas a dados intervalares, no contexto de regressão. Dados sintéticos e reais são utilizados para comparar os métodos provenientes das representações propostas e aqueles presentes na literatura. / An interval is a complex data type used in the information aggregation or in the representation of imprecise data. This work presents two new representations of intervals in order to construct a new cluster method and a new linear regression method for this kind of data. Dynamic clustering defines partitions into the data and it defines prototypes associated with each one of these partitions. The prototypes summarize the information about the partitions and they are used in a minimization criterion which depends on a distance, which is responsible for quantifying the proximity between instances and prototypes. In this way, it is proposed a new hybrid distance between intervals based on a family of distances between points. Points are obtained from the interval through a mapping. Also, it is proposed two versions of the hybrid distance, both with weights: one with weights in hybridism and other with adaptive weights. In linear regression, it is proposed to represent the intervals through the parametric equation of the line. This parametrization allows to find the set of points in the regression variables corresponding to the best estimates for the response variable limits. Before the regression construction, a criterion is computed to verify the mathematical consistency of prediction, where the upper limit must be greater than or equal to the lower. If the test shows that consistency is not guaranteed, then the application proposes a transformation of the response variable. Therefore, this work also proposes some transformations that can be applied to interval data in the regression context. Synthetic and real data are used to compare the proposed methods and those one proposed on literature.

Pattern recognition has as its objective to classify objects into different categories and classes. It is a fundamental component of artificial intelligence and computer vision. This thesis investigates the application of an efficient optimization method, known as Particle Swarm Optimization (PSO), to the field of pattern recognition and image processing. First a clustering method that is based on PSO is proposed. The application of the proposed clustering algorithm to the problem of unsupervised classification and segmentation of images is investigated. A new automatic image generation tool tailored specifically for the verification and comparison of various unsupervised image classification algorithms is then developed. A dynamic clustering algorithm which automatically determines the "optimum" number of clusters and simultaneously clusters the data set with minimal user interference is then developed. Finally, PSO-based approaches are proposed to tackle the color image quantization and spectral unmixing problems. In all the proposed approaches, the influence of PSO parameters on the performance of the proposed algorithms is evaluated. / Thesis (PhD)--University of Pretoria, 2006. / Computer Science / unrestricted

This project is about trying to apply machine learning theories on a selection of data points in order to see if an improvement of current methodology within stress detection and measure selecting could be applicable for the company Linkura AB. Linkura AB is a medical technology company based in Linköping and handles among other things stress measuring for different companies employees, as well as health coaching for selecting measures. In this report we experiment with different methods and algorithms under the collective name of Unsupervised Learning, to identify visible patterns and behaviour of data points and further on we analyze it with the quantity of data received. The methods that have been practiced on during the project are “K-means algorithm” and a dynamic hierarchical clustering algorithm. The correlation between the different data points parameters is analyzed to optimize the resource consumption, also experiments with different number of parameters are tested and discussed with an expert in stress coaching. The results stated that both algorithms can create clusters for the risk groups, however, the dynamic clustering method clearly demonstrate the optimal number of clusters that should be used. Having consulted with mentors and health coaches regarding the analysis of the produced clusters, a conclusion that the dynamic hierarchical cluster algorithm gives more accurate clusters to represent risk groups were done. The conclusion of this project is that the machine learning algorithms that have been used, can categorize data points with stress behavioral correlations, which is usable in measure testimonials. Further research should be done with a greater set of data for a more optimal result, where this project can form the basis for the implementations. / Detta projekt handlar om att försöka applicera maskininlärningsmodeller på ett urval av datapunkter för att ta reda på huruvida en förbättring av nuvarande praxis inom stressdetektering och  åtgärdshantering kan vara applicerbart för företaget Linkura AB. Linkura AB är ett medicintekniskt företag baserat i Linköping och hanterar bland annat stressmätning hos andra företags anställda, samt hälso-coachning för att ta fram åtgärdspunkter för förbättring. I denna rapport experimenterar vi med olika metoder under samlingsnamnet oövervakad maskininlärning för att identifiera synbara mönster och beteenden inom datapunkter, och vidare analyseras detta i förhållande till den mängden data vi fått tillgodosett. De modeller som har använts under projektets gång har varit “K-Means algoritm” samt en dynamisk hierarkisk klustermodell. Korrelationen mellan olika datapunktsparametrar analyseras för att optimera resurshantering, samt experimentering med olika antal parametrar inkluderade i datan testas och diskuteras med expertis inom hälso-coachning. Resultaten påvisade att båda algoritmerna kan generera kluster för riskgrupper, men där den dynamiska modellen tydligt påvisar antalet kluster som ska användas för optimalt resultat. Efter konsultering med mentorer samt expertis inom hälso-coachning så drogs en slutsats om att den dynamiska modellen levererar tydligare riskkluster för att representera riskgrupper för stress. Slutsatsen för projektet blev att maskininlärningsmodeller kan kategorisera datapunkter med stressrelaterade korrelationer, vilket är användbart för åtgärdsbestämmelser. Framtida arbeten bör göras med ett större mängd data för mer optimerade resultat, där detta projekt kan ses som en grund för dessa implementeringar.

Submitted by Izabel Franco (izabel-franco@ufscar.br) on 2016-09-06T18:59:08Z No. of bitstreams: 1 TeseFJJS.pdf: 3277696 bytes, checksum: 0a34a4499fb5e482fa95ea8925603968 (MD5) / Approved for entry into archive by Marina Freitas (marinapf@ufscar.br) on 2016-09-12T14:12:50Z (GMT) No. of bitstreams: 1 TeseFJJS.pdf: 3277696 bytes, checksum: 0a34a4499fb5e482fa95ea8925603968 (MD5) / Approved for entry into archive by Marina Freitas (marinapf@ufscar.br) on 2016-09-12T14:13:02Z (GMT) No. of bitstreams: 1 TeseFJJS.pdf: 3277696 bytes, checksum: 0a34a4499fb5e482fa95ea8925603968 (MD5) / Made available in DSpace on 2016-09-12T14:13:13Z (GMT). No. of bitstreams: 1 TeseFJJS.pdf: 3277696 bytes, checksum: 0a34a4499fb5e482fa95ea8925603968 (MD5) Previous issue date: 2015-04-30 / Não recebi financiamento / The good results obtained by the fuzzy approaches applied in the analysis of time series (TS) has contributed significantly to the growth of the area. Although there are satisfactory results in TS analysis with methods that use the classic concepts of TS and with the recent concepts of fuzzy time series (FTS), there is a lack of models combining both areas. Face of this context, the contributions of this thesis are associated with the development of models for TS analysis combining the concepts of FTS with statistical methods aiming at the improvement in accuracy of forecasts and in identification of behavioral changes in the TS. In order to allow a suitable fuzzy representation of crisp values observed, the approaches developed in this thesis were combined with a new proposal for pre-processing of the data. The prediction value is calculated from a new smoothing technique combined with an extension of the fuzzy logic relationships. This combination allow to be considered in value computed different degrees of influence to the most recent behavior and to the oldest behavior of the series. In situations where the model does not have the necessary knowledge to calculate the predicted value, the concepts of simple linear regression are combined with the concepts of the FTS to identify the most recent trend in the TS. The approach developed for the behavioral analysis of the TS aims to identify changes in behavior from the definition of prototypes that represent the groups of the TS and from the segmentation of the series that will be analyzed. In this new approach, the dissimilarity between a segment of a TS and the corresponding interval of a given prototype is defined by metric Fuzzy Dynamic Time Warping weighted by a new smoothing technique applied to the distance matrix between the observed data. The accuracy obtained by the forecast model not only demonstrates the effectiveness of the developed approach, but also shows the evolution of model throughout the research and the importance of preprocessing in the forecast. The analysis of segmented TS identifies satisfactorily the behavioral changes of the series by calculating the membership functions of these segments in the respective groups represented by the prototypes. / Os bons resultados obtidos pelas abordagens fuzzy utilizadas para a análise de séries temporais (ST) tem contribuído significativamente para o crescimento da área. Embora haja resultados satisfatórios na análise de ST com métodos que utilizam os conceitos clássicos de ST e também com os conceitos recentes de séries temporais fuzzy (STF), há uma carência de modelos que combinem ambas as áreas. Diante deste contexto, as contribuições deste trabalho estão associadas ao desenvolvimento de modelos para a análise de ST combinando os conceitos de STF e métodos estatísticos visando a melhora na acurácia das previsões e a identificação de alterações comportamentais nas séries. Com o objetivo de permitir uma melhor representação fuzzy dos valores crisp observados, as abordagens desenvolvidas nesta tese foram associadas a uma nova proposta de pré-processamento dos dados. A previsão de valores é calculada a partir de uma nova técnica de suavização combinada a uma extensão das relações lógicas fuzzy. Essa combinação permite que sejam considerados no cálculo do valor previsto diferentes graus de influência para o comportamento mais recente e para o comportamento mais antigo da série. Em ocasiões onde o modelo não dispõe do conhecimento necessário para o cálculo do valor previsto, os conceitos de regressão linear simples são associados aos conceitos das STF para identificar a tendência mais recente da ST. A abordagem desenvolvida para a análise comportamental das séries tem como objetivo identificar mudanças no comportamento a partir da definição de protótipos que representam um grupo de ST e da segmentação das séries a serem analisadas. Nesta nova abordagem, a dissimilaridade entre um segmento de uma ST e o intervalo correspondente de um determinado protótipo é definida por meio da métrica Dynamic Time Warping (DTW) Fuzzy, ponderada por uma nova técnica de suavização aplicada à matriz de distâncias entre os dados observados. A acurácia obtida pelo modelo de previsão não só comprova a eficácia da abordagem desenvolvida, como também demonstra a evolução do modelo ao longo da pesquisa e a importância do pré-processamento nas previsões. A análise das ST segmentadas identifica satisfatoriamente as alterações comportamentais das séries por meio do cálculo da pertinência dos segmentos nos respectivos grupos representados pelos protótipos.

Les réseaux de capteurs sans fil est une technologie nouvelle dont les applications s'étendent sur plusieurs domaines: militaire, scientifique, médicale, industriel, etc. La collaboration entre les noeuds capteurs, caractérisés par des capacités minimales en termes de capture, de transmission, de traitement et d'énergie, est une nécessité pour réaliser des tâches aussi complexes que la collecte des données, le pistage des objets mobiles, la surveillance des zones sensibles, etc. La contrainte matérielle sur le développement des ressources énergétiques des noeuds capteurs est persistante. D'où la nécessité de l'optimisation logicielle dans les différentes couches de la pile protocolaire et du système d'exploitation des noeuds. Dans cette thèse, nous approchons le problème d'optimisation d'énergie pour les applications collaboratives via les méthodes de sélection des capteurs basées sur la prédiction et la corrélation des données issues du réseau lui-même. Nous élaborons plusieurs méthodes pour conserver les ressources énergétiques du réseau en utilisant la prédiction comme un moyen pour anticiper les actions des noeuds et leurs rôles afin de minimiser le nombre des noeuds impliqués dans la tâche en question. Nous prenons l'application de pistage d'objets mobiles comme un cas d'étude. Ceci, après avoir dresser un état de l'art des différentes méthodes et approches récentes utilisées dans ce contexte. Nous formalisons le problème à l'aide d'un programme linéaire à variables binaires dans le but de trouver une solution générale exacte. Nous modélisons ainsi le problème de minimisation de la consommation d'énergie des réseaux de capteurs sans fil, déployé pour des applications de collecte de données soumis à la contrainte de précision de données, appelé EMDP. Nous montrons que ce problème est NP-Complet. D'où la nécessité de solutions heuristiques. Comme solution approchée, nous proposons un algorithme de clustering dynamique, appelé CORAD, qui adapte la topologie du réseau à la dynamique des données capturées afin d'optimiser la consommation d'énergie en exploitant la corrélation qui pourrait exister entre les noeuds. Toutes ces méthodes ont été implémentées et testées via des simulations afin de montrer leur efficacité. / Wireless Sensor Networks is an emerging technology enabled by the recent advances in Micro-Electro-Mechanical Systems, that led to design tiny wireless sensor nodes characterized by small capacities of sensing, data processing and communication. To accomplish complex tasks such as target tracking, data collection and zone surveillance, these nodes need to collaborate between each others to overcome the lack of battery capacity. Since the development of the batteries hardware is very slow, the optimization effort should be inevitably focused on the software layers of the protocol stack of the nodes and their operating systems. In this thesis, we investigated the energy problem in the context of collaborative applications and proposed an approach based on node selection using predictions and data correlations, to meet the application requirements in terms of energy-efficiency and quality of data. First, we surveyed almost all the recent approaches proposed in the literature that treat the problem of energy-efficiency of prediction-based target tracking schemes, in order to extract the relevant recommendations. Next, we proposed a dynamic clustering protocol based on an enhanced version of the Distributed Kalman Filter used as a prediction algorithm, to design an energy-efficient target tracking scheme. Our proposed scheme use these predictions to anticipate the actions of the nodes and their roles to minimize their number in the tasks. Based on our findings issued from the simulation data, we generalized our approach to any data collection scheme that uses a geographic-based clustering algorithm. We formulated the problem of energy minimization under data precision constraints using a binary integer linear program to find its exact solution in the general context. We validated the model and proved some of its fundamental properties. Finally and given the complexity of the problem, we proposed and evaluated a heuristic solution consisting of a correlation-based adaptive clustering algorithm for data collection. We showed that, by relaxing some constraints of the problem, our heuristic solution achieves an acceptable level of energy-efficiency while preserving the quality of data.