[ES] La sectorización en sectores hidrométricos de las redes de abastecimiento de agua potable ha sido, a lo largo de los últimos años, una de las técnicas más empleadas para ubicar y controlar las fugas. Esto es de vital importancia ante un recurso, el agua potable, cada vez más escaso y valioso. Sin embargo, el empleo de esta técnica, algunas veces, es llevada a cabo por ensayo y error, o sin una estrategia previamente establecida. La capacidad hidráulica de la red, o caudal máximo que puede aportar cumpliendo los requisitos mínimos de servicio, puede afectarse cuando se lleva a cabo un proceso de sectorización. Esto es debido a que el cierre de tramos de la red cambia las condiciones de servicio, pudiendo generar afectaciones graves desde el punto de vista de presiones y de calidad de agua. Algunas veces, también, es aplicado el proceso de sectorización sin una evaluación previa de las características de la red y de las condiciones operativas. Partiendo de estas situaciones, se planteó indagar acerca de cómo las condiciones iniciales de operación de la red pudiesen afectar la formación de los sectores hidrométricos desde el punto de vista de su configuración. En los procesos de sectorización pueden emplearse algoritmos de clústering debido a la analogía existente entre los agrupamientos de los diferentes elementos de los sistemas de abastecimiento de agua y los principios aplicados en las técnicas de agrupamiento de datos. Así, se decidió aplicar diversas técnicas basadas en la detección de comunidades en redes, ya que poseen la ventaja de agrupar en función de la información en las conexiones, y no solo en la información almacenada en los nodos, lo cual resulta favorable para el tema en estudio en este trabajo, debido a que pudiera tomarse en cuenta cómo la variación de los niveles de exigencia en las tuberías, ocasionados por la variación en la demanda, afectaría la formación de sectores. Para ello se planteó trabajar con dos redes de prueba conocidas en la literatura, a las cuales se les simuló bajo distintas capacidades de carga de trabajo y a través del empleo de varios algoritmos de agrupamiento, tales como Infomap, Walktrap, Fast gready, Edge betweeness, Luovain y Leiden. La metodología desarrollada fue, finalmente, aplicada a una red real de gran tamaño, lo que valida dicha metodología para un problema del mundo real. Como parámetros para ponderar las conexiones se probaron cuatro opciones: caudal circulante con la red trabajando a la capacidad deseada; energía disipada en cada tramo con la red trabajando a la capacidad deseada; relación entre el caudal circulante en el tramo y el caudal circulante a capacidad máxima teórica de la red; y relación entre la energía disipada trabajando a la capacidad deseada y la disipación de la energía en los tramos con la red trabajando a capacidad máxima. Obtenida la partición por cualquiera de los métodos anteriormente expuestos, se optimizaron las entradas y salidas de los sectores por medio de la implementación de un algoritmo de enjambre de partículas; adicionalmente, se hace una comparación del resultado con otro obtenido a partir de un algoritmo de colonia de hormigas, lo que permite dar robustez a la metodología desarrollada. / [CA] La partició o sectorització de les xarxes d'abastiment d'aigua potable a les anomenades districtes hidromètrics ha estat, durant els darrers anys, una de les tècniques més utilitzades per localitzar i controlar les pèrdues d'aigua. La sectorització és crucial quan es tracta d'un recurs cada vegada més escàs i valuós com es l'aigua. Tanmateix, l'ús d'aquesta tècnica de vegades es duu a terme mitjançant proves i errors, o sense una estratègia prèviament establerta. La capacitat hidràulica d'una xarxa o el cabal màxim que pot proporcionar mentre compleix els requisits mínims de servei es pot veure afectat quan es realitza un procés de sectorització. Això es deu al fet que el tancament de seccions de la xarxa modifica les condicions del servei, la qual cosa pot generar greus efectes des del punt de vista de la pressió i la qualitat de l'aigua. A més, de vegades s'apliquen processos de sectorització sense una avaluació prèvia de les característiques de la xarxa i de les condicions de funcionament. Partint d'aquestes situacions, ens vam preguntar sobre com les condicions inicials de funcionament de la xarxa podrien afectar la construcció dels sectors hidromètrics des del punt de vista de la seva configuració. Els algoritmes de clusterització es poden utilitzar en processos de sectorització a causa de l'analogia entre l'agrupació dels diferents elements dels sistemes d'abastiment d'aigua i els principis aplicats en les tècniques d'agrupament de dades. Així, vam decidir aplicar diverses tècniques basades en la detecció de comunitats a les xarxes, ja que tenen l'avantatge de realitzar agrupacions basades en la informació de les unions i no només en la informació emmagatzemada als nodes. Això és favorable per al tema en estudi en aquest treball, perquè es pot considerar com la variació dels nivells de demanda a les canonades, causada per la variació de la demanda, pot afectar la construcció de sectors. Amb aquesta finalitat, es va proposar treballar amb dues xarxes de proves conegudes a la literatura, que es van simular sota diferents capacitats de càrrega i mitjançant l'ús de diversos algorismes de clusterització, com Infomap, Walktrap, Fast gready, Edge betweeness, Luovain i Leiden. La metodologia desenvolupada es va aplicar, finalment, a una gran xarxa real, que valida la metodologia per a una xarxa de subministrament d'aigua del món real. Com a paràmetres per ponderar les unions, es van provar quatre opcions: cabal circulant amb la xarxa treballant a la capacitat desitjada; energia dissipada a cada secció amb la xarxa treballant a la capacitat desitjada; relació entre el cabal circulant a la secció i el cabal circulant a la capacitat màxima teòrica de la xarxa; i la relació entre l'energia dissipada que treballa a la capacitat desitjada i la dissipació d'energia a les seccions amb la xarxa que treballa a la màxima capacitat. Un cop obtinguda la partició per qualsevol dels mètodes esmentats anteriorment, les entrades i sortides dels sectors s'optimitzen mitjançant un algorisme d'optimització d'eixam de partícules; a més, es fa una comparació del resultat amb un altre algoritme d'optimització, és a dir, un algoritme d'optimització de colònies de formigues, que permet proporcionar robustesa a la metodologia desenvolupada. / [EN] Partitioning or sectorization of drinking water supply networks into so-called district metered areas has been, over the last years, one of the most used techniques to locate and control leaks. Sectorization is crucial when dealing with an increasingly scarce and valuable resource as water. However, the use of this technique is sometimes carried out by trial and error, or without a previously established strategy. The hydraulic capacity of a network, or the maximum flow it can provide while meeting the minimum service requirements, can be affected when a sectorization process is carried out. This is due to the fact that the closure of sections of the network changes the service conditions, which could generate serious effects from the point of view of pressure and water quality. Also, sometimes sectorization processes are applied without a prior evaluation of the characteristics of the network and of the operating conditions. Starting from these situations, we inquired about how the initial operating conditions of the network could affect the construction of the hydrometric sectors or district metered areas from the point of view of their configuration. Clustering algorithms can be used in sectorization processes due to the analogy between the grouping of the different elements of water supply systems and the principles applied in data grouping techniques. Thus, we decided to apply various techniques based on the detection of communities in networks, since they have the advantage of performing grouping based on the information at the junctions, and not only on the information stored at the nodes. This is favourable for the subject under study in this work, because one may consider how the variation of the demand levels in the pipes, caused by the demand variation, can affect the construction of sectors. To this end, it was proposed to work with two test networks known in the literature, which were simulated under different load capacities and through the use of various clustering algorithms, namely Infomap, Walktrap, Fast gready, Edge betweeness, Luovain and Leiden. The methodology developed was, finally, applied to a large real network, which validates the methodology for a real-world water supply network. As parameters to weight the junctions, four options were tested: circulating flow with the network working at the desired capacity; energy dissipated in each section with the network working at the desired capacity; relationship between the circulating flow in the section and the circulating flow at the theoretical maximum capacity of the network; and the relationship between the energy dissipated working at the desired capacity and the dissipation of energy in the sections with the network working at maximum capacity. Once the partition is obtained by any of the previously said methods, the inputs and outputs of the sectors are optimized by using a particle swarm optimization algorithm; additionally, a comparison of the result is made with another optimization algorithm, namely, an ant colony optimization algorithm, which allows to provide robustness to the developed methodology. / Martins Alves, C. (2020). La capacidad hidráulica en las redes de agua potable y su influencia en el proceso de sectorización [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/159938 / TESIS

In this dissertation we aim to perform a multi-wavelength analysis of galaxy clusters. We discuss various methods for clustering in order to determine physical parameters of galaxy clusters required for this type of study. A selection of galaxy clusters was chosen from 4 papers, (Popesso et al. 2007b, Yoon et al. 2008, Loubser et al. 2008, Brownstein & Mo at 2006) and restricted by redshift and galactic latitude to reveal a sample of 40 galaxy clusters with 0.0 < z < 0.15. Data mining using Virtual Observatory (VO) and a literature survey provided some background information about each of the galaxy clusters in our sample with respect to optical, radio and X-ray data. Using the Kayes Mixture Model (KMM) and the Gaussian Mixing Model (GMM), we determine the most likely cluster member candidates for each source in our sample. We compare the results obtained to SIMBADs method of hierarchy. We show that the GMM provides a very robust method to determine member candidates but in order to ensure that the right candidates are chosen we apply a select choice of outlier tests to our sources. We determine a method based on a combination of GMM, the QQ Plot and the Rosner test that provides a robust and consistent method for determining galaxy cluster members. Comparison between calculated physical parameters; velocity dispersion, radius, mass and temperature, and values obtained from literature show that for the majority of our galaxy clusters agree within 3 range. Inconsistencies are thought to be due to dynamically active clusters that have substructure or are undergoing mergers, making galaxy member identi cation di cult. Six correlations between di erent physical parameters in the optical and X-ray wavelength were consistent with published results. Comparing the velocity dispersion with the X-ray temperature, we found a relation of T0:43 as compared to T0:5 obtained from Bird et al. (1995). X-ray luminosity temperature and X-ray luminosity velocity dispersion relations gave the results LX T2:44 and LX 2:40 which lie within the uncertainty of results given by Rozgacheva & Kuvshinova (2010). These results all suggest that our method for determining galaxy cluster members is e cient and application to higher redshift sources can be considered. Further studies on galaxy clusters with substructure must be performed in order to improve this method. In future work, the physical parameters obtained here will be further compared to X-ray and radio properties in order to determine a link between bent radio sources and the galaxy cluster environment. / MSc (Space Physics), North-West University, Potchefstroom Campus, 2013

In this dissertation we aim to perform a multi-wavelength analysis of galaxy clusters. We discuss various methods for clustering in order to determine physical parameters of galaxy clusters required for this type of study. A selection of galaxy clusters was chosen from 4 papers, (Popesso et al. 2007b, Yoon et al. 2008, Loubser et al. 2008, Brownstein & Mo at 2006) and restricted by redshift and galactic latitude to reveal a sample of 40 galaxy clusters with 0.0 < z < 0.15. Data mining using Virtual Observatory (VO) and a literature survey provided some background information about each of the galaxy clusters in our sample with respect to optical, radio and X-ray data. Using the Kayes Mixture Model (KMM) and the Gaussian Mixing Model (GMM), we determine the most likely cluster member candidates for each source in our sample. We compare the results obtained to SIMBADs method of hierarchy. We show that the GMM provides a very robust method to determine member candidates but in order to ensure that the right candidates are chosen we apply a select choice of outlier tests to our sources. We determine a method based on a combination of GMM, the QQ Plot and the Rosner test that provides a robust and consistent method for determining galaxy cluster members. Comparison between calculated physical parameters; velocity dispersion, radius, mass and temperature, and values obtained from literature show that for the majority of our galaxy clusters agree within 3 range. Inconsistencies are thought to be due to dynamically active clusters that have substructure or are undergoing mergers, making galaxy member identi cation di cult. Six correlations between di erent physical parameters in the optical and X-ray wavelength were consistent with published results. Comparing the velocity dispersion with the X-ray temperature, we found a relation of T0:43 as compared to T0:5 obtained from Bird et al. (1995). X-ray luminosity temperature and X-ray luminosity velocity dispersion relations gave the results LX T2:44 and LX 2:40 which lie within the uncertainty of results given by Rozgacheva & Kuvshinova (2010). These results all suggest that our method for determining galaxy cluster members is e cient and application to higher redshift sources can be considered. Further studies on galaxy clusters with substructure must be performed in order to improve this method. In future work, the physical parameters obtained here will be further compared to X-ray and radio properties in order to determine a link between bent radio sources and the galaxy cluster environment. / MSc (Space Physics), North-West University, Potchefstroom Campus, 2013

La Science des réseaux est apparue comme un domaine d'étude fondamental pour modéliser un grand nombre de systèmes synthétiques ou du monde réel.La découverte du graphe petit monde et du graphe sans échelle dans ces réseaux a révolutionné la façon d'étudier, d'analyser, de modéliser et de traiter ces réseaux. Dans cette thèse, nous nous intéressons à l'étude des réseaux ayant ces propriétés et souvent qualifiés de réseaux complexes.A notre avis, les recherches menées dans ce domaine peuvent être regroupées en quatre catégories: l'analyse, la structure, le processus/organisation et la visualisation.Nous abordons des problèmes relatifs à chacune de ces catégories tout au long de cette thèse. (...) / Network science has emerged as a fundamental field of study to model many physicaland real world systems around us. The discovery of small world and scale free propertiesof these real world networks has revolutionized the way we study, analyze, model andprocess these networks. In this thesis, we are interested in the study of networks havingthese properties often termed as complex networks. In our opinion, research conducted inthis field can be grouped into four categories, Analysis, Structure, Processes-Organizationand Visualization. We address problems pertaining to each of these categories throughoutthis thesis. (...)

Power limits on future high-performance computing (HPC) systems will constrain applications. However, HPC applications do not consume constant power over their lifetimes. Thus, applications assigned a fixed power bound may be forced to slow down during high-power computation phases, but may not consume their full power allocation during low-power I/O phases. This paper explores algorithms that leverage application semantics-phase frequency, duration and power needs-to shift unused power from applications in I/O phases to applications in computation phases, thus improving system-wide performance. We design novel techniques that include explicit staggering of applications to improve power shifting. Compared to executing without power shifting, our algorithms can improve average performance by up to 8% or improve performance of a single, high-priority application by up to 32%.

Submitted by Johnny Rodrigues (johnnyrodrigues@ufcg.edu.br) on 2018-08-04T13:59:37Z No. of bitstreams: 1 MAXWELL GUIMARÃES DE OLIVEIRA - DISSERTAÇÃO PPGCC 2012..pdf: 28277196 bytes, checksum: 398cd7b385ee61c414d0086810fbeeed (MD5) / Made available in DSpace on 2018-08-04T13:59:37Z (GMT). No. of bitstreams: 1 MAXWELL GUIMARÃES DE OLIVEIRA - DISSERTAÇÃO PPGCC 2012..pdf: 28277196 bytes, checksum: 398cd7b385ee61c414d0086810fbeeed (MD5) Previous issue date: 2012-08-20 / Capes / Atualmente, há um volume considerável de dados espaço-temporais disponíveis em vários meios, sobretudo na Internet. A visualização de dados espaço-temporais é uma tarefa complexa, que requer uma série de recursos visuais apropriados para que, em conjunto, possam permitir aos usuários uma correta interpretação das informações analisadas. Além do emprego de técnicas de visualização, a utilização de técnicas de descoberta de conhecimento em bancos de dados tem se mostrado relevante no auxílio à análise exploratória de relacionamentos em dados espaço-temporais. O levantamento do estado da arte em visualização de dados espaço-temporais leva à conclusão de que a área ainda é deficiente em soluções para visualização e análise desses tipos. Muitas abordagens abrangem somente questões espaciais, desprezando as características temporais desses dados. Inserido nesse contexto, o principal objetivo deste trabalho é melhorar a experiência do usuário em visualização e análise espaço-temporal, indo além do universo da visualização dos dados espaço-temporais brutos e considerando, também, a importância em visualização de dados espaço-temporais derivados de um processo de descoberta de conhecimento, mais especificamente algoritmos de clustering. Esse objetivo é atingido com a definição de uma abordagem inovadora em visualização de dados espaço-temporais, e de sua implementação, denominada GeoSTAT (Geographic SpatioTemporal Analysis Tool), que engloba pontos importantes observados nas principais abordagens existentes e acrescenta, principalmente, técnicas de visualização voltadas à dimensão temporal e à utilização de algoritmos de clustering, valorizando características até então pouco exploradas em dados espaço-temporais. A validação deste trabalho ocorre por meio de dois estudos de caso, onde cada um aborda dados espaço-temporais de um domínio específico, para demonstrar a experiência do usuário final diante das técnicas de visualização reunidas na abordagem proposta. / Nowadays, there is a considerable amount of spatiotemporal data available in various media, especially on the Internet. The visualization of spatiotemporal data is a complex task that requires a series of visual suitable resources which can enable users to have a correct interpretation of the data. Apart from the use of visualization techniques, the use of techniques of knowledge discovery in databases has proven relevantfor the exploratory analysis of relationships in spatiotemporal data. The state of the art in visualization of spatiotemporal data leads to the conclusion that the area is still deficient in solutions for viewing and analysis of those data. Many approaches cover only spatial issues, ignoring the temporal characteristics of such data. Inserted in this context, the main objective of this work is to improve the user experience in spatiotemporal visualization and analysis, going beyond the universe of visualization of spatiotemporal raw data and also considering the importance of visualization of spatiotemporal data derived from a knowledge discovery process, more specifically clustering algorithms. This goal is achieved by defining an innovative approach for the analysis and visualization of spatiotemporal data, and its implementation, called GeoSTAT (Spatiotemporal Geographic Analysis Tool), which includes importam points in the main existing approaches and adds especially visualization techniques geared to the temporal dimension and the use of clustering algorithms, enhancing unexplored features in spatiotemporal data. The validation of this work occurs through two case studies, where each one deals with spatiotemporal data of a specific domain to demonstrate the end-user experience on the visualization techniques combined in the proposed approach.

En bio-informatique, comprendre comment les molécules d’ADN ont évolué au cours du temps reste un problème ouvert etcomplexe. Des algorithmes ont été proposés pour résoudre ce problème, mais ils se limitent soit à l’évolution d’un caractèredonné (par exemple, un nucléotide précis), ou se focalisent a contrario sur de gros génomes nucléaires (plusieurs milliardsde paires de base), ces derniers ayant connus de multiples événements de recombinaison – le problème étant NP completquand on considère l’ensemble de toutes les opérations possibles sur ces séquences, aucune solution n’existe à l’heureactuelle. Dans cette thèse, nous nous attaquons au problème de reconstruction des séquences ADN ancestrales en nousfocalisant sur des chaînes nucléotidiques de taille intermédiaire, et ayant connu assez peu de recombinaison au coursdu temps : les génomes de chloroplastes. Nous montrons qu’à cette échelle le problème de la reconstruction d’ancêtrespeut être résolu, même quand on considère l’ensemble de tous les génomes chloroplastiques complets actuellementdisponibles. Nous nous concentrons plus précisément sur l’ordre et le contenu ancestral en gènes, ainsi que sur lesproblèmes techniques que cette reconstruction soulève dans le cas des chloroplastes. Nous montrons comment obtenirune prédiction des séquences codantes d’une qualité telle qu’elle permette ladite reconstruction, puis comment obtenir unarbre phylogénétique en accord avec le plus grand nombre possible de gènes, sur lesquels nous pouvons ensuite appuyernotre remontée dans le temps – cette dernière étant en cours de finalisation. Ces méthodes, combinant l’utilisation d’outilsdéjà disponibles (dont la qualité a été évaluée) à du calcul haute performance, de l’intelligence artificielle et de la biostatistique,ont été appliquées à une collection de plus de 450 génomes chloroplastiques. / In Bioinformatics, understanding how DNA molecules have evolved over time remains an open and complex problem.Algorithms have been proposed to solve this problem, but they are limited either to the evolution of a given character (forexample, a specific nucleotide), or conversely focus on large nuclear genomes (several billion base pairs ), the latter havingknown multiple recombination events - the problem is NP complete when you consider the set of all possible operationson these sequences, no solution exists at present. In this thesis, we tackle the problem of reconstruction of ancestral DNAsequences by focusing on the nucleotide chains of intermediate size, and have experienced relatively little recombinationover time: chloroplast genomes. We show that at this level the problem of the reconstruction of ancestors can be resolved,even when you consider the set of all complete chloroplast genomes currently available. We focus specifically on the orderand ancestral gene content, as well as the technical problems this raises reconstruction in the case of chloroplasts. Weshow how to obtain a prediction of the coding sequences of a quality such as to allow said reconstruction and how toobtain a phylogenetic tree in agreement with the largest number of genes, on which we can then support our back in time- the latter being finalized. These methods, combining the use of tools already available (the quality of which has beenassessed) in high performance computing, artificial intelligence and bio-statistics were applied to a collection of more than450 chloroplast genomes.

No description available.

The thesis deals with text mining. It describes the theory of text document clustering as well as algorithms used for clustering. This theory serves as a basis for developing an application for clustering text data. The application is developed in Java programming language and contains three methods used for clustering. The user can choose which method will be used for clustering the collection of documents. The implemented methods are K medoids, BiSec K medoids, and SOM (self-organization maps). The application also includes a validation set, which was specially created for the diploma thesis and it is used for testing the algorithms. Finally, the algorithms are compared according to obtained results.

This diploma thesis deals with the use of some methods of cluster analysis on the ECG signal in order to sort QRS complexes according to their morphology to normal and abnormal. It is used agglomerative hierarchical clustering and non-hierarchical method K – Means for which an application in Mathworks MATLAB programming equipment was developed. The first part deals with the theory of the ECG signal and cluster analysis, and then the second is the design, implementation and evaluation of the results of the usage of developed software on the ECG signal for the automatic division of QRS complexes into clusters.