Ossos, estruturas essenciais para a proteção de órgãos internos, estrutura corporal e suporte mecânico nos vertebrados, possuem uma complexa rede de canais (canais de Volkmann e Havers) responsáveis por nutrir as células do tecido. Entretanto a falta de estudos quantitativos leva a uma carência de medidas e parâmetros para a caracterização dessas estruturas. Utilizando computação gráfica, técnicas de processamento de imagens, e redes complexas descreveremos a obtenção, reconstrução, representação, e análise dessas redes de canais. Para isso, duas falanges distais, uma de um galo e uma de uma galinha, passaram por um processo de corte histológico, as lâminas resultantes foram fotografadas e as imagens tratadas até serem reconstruídas em 3D. Os volumes foram convertidos em redes complexas, o que permitiu o uso de métodos de análise consagrados pela literatura. As redes foram comparadas entre si e com a rede do trabalho desenvolvido por Matheus P. Viana et al. (1–3) usando análise de grau, posicionamento dos nós, detecção de comunidades, e ataques (em cascata e aleatórios). Três resultados se destacam: 1) as redes apresentam diviões predominantemente dicotômica dos canais; 2) as redes apresentam uma alta modularidade, indicando que áreas específicas desempenham funções específicas; e 3) as redes são particularmente resistentes a ataques em cascata. / Bones are essential for the protection of internal organs, for body structure, and for mechanical support in vertebrates, and present a complex network of channels (Havers and Volkmann channels) required to nourish tissue cells. However, the lack of quantitative studies leads to scarce parameters and measures to characterize these structures. By using computational graphic, image processing, and complex networks we will describe the acquisitation, reconstruction, representation, and analysis of these channel networks. Two distal phalanges (one from a hen and one from a rooster) were submitted to hystological section processing; the resulting slices were photographed and the images were treated before 3D reconstruction. The volumes were converted into complex networks which allow us to use methods of analysis widely accepted in literature. Networks were compared with each other and with the network obtained in the study by Viana et al. (1–3) using degree analysis, node positioning, community detection, and random and systematic attacks. Three results stand out: (i) the networks show a predominantly dichotomic division of channels; (ii) the networks show high modularity, indicating that specific areas perform specific functions; and (iii) the networks are particularly resistant to cascate attacks.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2017-05-31T12:29:07Z No. of bitstreams: 1 abraaoguimaraesflores.pdf: 5791600 bytes, checksum: 384ace33f746857745754f651589590f (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2017-05-31T12:42:46Z (GMT) No. of bitstreams: 1 abraaoguimaraesflores.pdf: 5791600 bytes, checksum: 384ace33f746857745754f651589590f (MD5) / Made available in DSpace on 2017-05-31T12:42:46Z (GMT). No. of bitstreams: 1 abraaoguimaraesflores.pdf: 5791600 bytes, checksum: 384ace33f746857745754f651589590f (MD5) Previous issue date: 2013-08-26 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / A dinamicidade é uma característica presente em diversos sistemas reais, tais como redes de comunicação, sociais, biológicas e tecnológicas. Processos de difusão em redes complexas podem surgir, por exemplo, em busca de dados, roteamento de dados e propa gação de doenças. Desta forma, a compreensão do tempo necessário para difusão é um tema de estudo importante em redes complexas dinâmicas. Nesta dissertação é realizado um estudo de como medidas de centralidade podem ajudar na diminuição do tempo de difusão de informação em redes complexas dinâmicas. Usando dados de sistemas reais e sintéticos é mostrado que, se a dinamicidade é desconsiderada, o tempo necessário para difundir uma informação na rede é subestimado. Foram propostos algoritmos de difusão que consideram métricas de centralidade em grafos. Estes algoritmos aceleram o processo de difusão, quando comparados com algoritmos de difusão mais simples, como o Random Walk. Por fim, foi analisado o impacto de um modelo simples de predição de arestas nos algoritmos de difusão baseados em centralidade que foram propostos nesta dissertação. / The dynamics is a characteristic present in many real systems, such as communication networks, social, biological and technological. Diffusion processes in complex networks may arise, for example, search data, routing data and the spread of diseases. Thus, understanding the time required for diffusion is an important topic of study in dynamic complex networks. This dissertation is a study of how centrality measures can help in reducing the time information dissemination in dynamic complex networks. Using data from synthetic and real systems is shown that if the dynamics is disregarded, the time needed for spreading an information network is underestimated. Diffusion algorithms have been proposed that consider metrics of centrality in graphs. Finally, we analyze the impact of a simple model for predicting edge algorithms in diffusion based on centrality that have been proposed in this dissertation.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2017-05-30T20:25:10Z No. of bitstreams: 1 franciscohenriquecerdeiraferreira.pdf: 1810078 bytes, checksum: b5d9796bdd636be5e4f0f3212220ad0a (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2017-06-01T11:36:20Z (GMT) No. of bitstreams: 1 franciscohenriquecerdeiraferreira.pdf: 1810078 bytes, checksum: b5d9796bdd636be5e4f0f3212220ad0a (MD5) / Made available in DSpace on 2017-06-01T11:36:20Z (GMT). No. of bitstreams: 1 franciscohenriquecerdeiraferreira.pdf: 1810078 bytes, checksum: b5d9796bdd636be5e4f0f3212220ad0a (MD5) Previous issue date: 2013-03-11 / Nos últimos anos, as aplicações P2P de transmissão de vídeo ao vivo despertaram um grande interesse na comunidade cientí ca. Essas aplicações geram um grande volume de dados que afetam diretamente o desempenho da rede. Apesar de existir um grande número de trabalhos dedicados a entender as aplicações P2P de transmissão de vídeo ao vivo, a maioria deles se baseia em uma visão estática desses sistemas. Estes trabalhos não se preocupam em entender a dinâmica do sistema, ou seja, como eles evoluem ao longo do tempo. Através de experimentos realizados na plataforma PlanetLab, este trabalho apresenta uma caracterização da aplicação SopCast, um dos mais importantes sistemas P2P de transmissão de vídeo ao vivo. O estudo é realizado levando-se em consideração a dinamicidade da topologia construída. Também é investigada a formação de comunidades na rede sobreposta e a correlação dessas comunidades com os Sistemas Autônomos da Internet (AS). Os resultados mostram que a formação de comunidades é bem de nida pela troca de tráfego e uma pequena porção de participantes desses grupos é responsável por sustentar toda a comunidade. Além disso, não existe indícios que os participantes se agrupam de acordo com seus Sistemas Autônomos. De fato, a probabilidade de uma comunidade ser formada com mais da metade de membros pertencentes a um mesmo AS é inferior a 10%. Finalmente, as caracterizações apresentadas fornecem informações importantes para o desenvolvimento de novas aplicações P2P de transmissão de vídeo ao vivo. Mais que isso, será possível algoritimos de formação de parcerias e grupos de tal forma que seja possível reduzir os custos de transmissão entre redes diferentes. / P2P live streaming systems have attracted a lot of attention from the research community in the last years. Such systems generate a large amount of data which impacts the network performance. Despite large number of works devoted to understand P2P live streaming applications, most of them rely on characterizing the static view of these systems. These works do not worry about either understanding the systems dynamics or analyzing how it evolves over time. Through experiments in PlanetLab platform, we present a characterization of the SopCast application, one of the most important P2P live streaming system. We focus on characterizing the dynamics of P2P overlay topology. We also investigated the community formation process in SopCast overlay and the correlation of these communities with the Autonomous Systems of the Internet (AS). Our results show that the formation of communities is well de ned by the exchange of tra c and a small portion of peers of these groups is responsible for supporting the entire community. Furthermore, there is a evidence that peeers are not grouped according to their Autonomous Systems. Indeed, the probability of a community be formed with more 50% of members belonging to the same AS is less than 10%. Finally, the characterizations we conduct provide important information to new P2P live streaming protocols and membership algorithms design. Moreover, the presented characterization may be useful to developers create algorithms that reduces the transmission cost on the P2P network.

Orientadores: José Antônio Brum, Marcus Aloizio Martinez de Aguiar / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Física Gleb Wataghin / Made available in DSpace on 2018-08-26T08:49:23Z (GMT). No. of bitstreams: 1 PinheiroNeto_Joao_M.pdf: 16027257 bytes, checksum: ff0a08fdba027bd6f2bc28198e34d482 (MD5) Previous issue date: 2014 / Resumo: Nesta dissertação de mestrado estudamos a dinâmica e estrutura de redes complexas. Começamos com uma revisão da literatura de redes complexas, apresentando as métricas de rede e modelos de conectividade mais comuns. Estudamos então em detalhe a dinâmica do modelo das Random Threshold Networks (RTN). Desenvolvemos uma nova aproximação de campo médio para a dinâmica de RTNs, consideravelmente mais simples que aproximações anteriores. Esta nova aproximação é útil de um ponto de vista prático, pois permite a geração de RTNs onde a atividade média da rede é controlável. Fazemos então uma revisão da literatura de redes adaptativas, apresentando alguns modelos de redes adaptativas com características interessantes. Por fim, desenvolvemos dois modelos de redes adaptativas inspirados pela evolução da estrutura neuronal no cérebro. O primeiro modelo utiliza regras simples e uma evolução baseada na remoção de links para controlar a atividade sobre a rede. A inspiração é a remoção de neurônios e conexões neuronais após a infância. Este modelo também consegue controlar a atividade de grupos individuais dentro de uma mesma rede. Exploramos uma variante deste modelo em um espaço bidimensional, onde conseguimos gerar redes modulares e small-world. O segundo modelo utiliza inputs externos para controlar a evolução da topologia da rede. A inspiração neste caso é o desenvolvimento das conexões neuronais durante a infância, que é influenciado por interações com o ambiente. O modelo gera avalanches finitas de atividade, e é capaz de gerar topologias especificas e modulares utilizando regras simples / Abstract: In this Masters Dissertation we study the structure and dynamics of complex networks. We start with a revision of the literature of complex networks, presenting the most common network metrics and models of network connectivity. We then study in detail the dynamics of the Random Threshold Network (RTN) model. We develop a new mean-field approximation for the RTN dynamics that is considerably more simple than previous results. This new approximation is useful from a practical standpoint, since it allows the generation of RTNs where the average activity of the network is controlled. We then review the literature of Adaptive Networks, explaining some of the adaptive models with interesting characteristics. At last, we develop two models of adaptive networks inspired by the evolution of neuronal structure in the brain. The first model uses simple rules and a link-removing evolution to control the activity on the network. The inspiration is the removal of neurons and neuronal connections after infancy. This model can also control the activity of individual groups within the same network. We explore a variant of this model in a bi-dimensional space, where we are able to generate modular and small-world networks. The second model uses external inputs to control the topological evolution of the network. The inspiration in this case is the development of neuronal connections during the infancy, which is influenced by interactions with the environment. The model generates finite avalanches of activity, and is capable of generating specific and modular topologies using simple rules / Mestrado / Física / Mestre em Física

A backbone network is the central part of the communication network, which provides connectivity within the various systems across large distances. Disruptions in a backbone network would cause severe consequences which could manifest in the service outage on a large scale. Depending on the size and the importance of the network, its failure could leave a substantial impact on the area it is associated with. The failures of the network services could lead to a significant disturbance of human activities. Therefore, making backbone communication networks more resilient directly affects the resilience of the area. Contemporary urban and regional development overwhelmingly converges with the communication infrastructure expansion and their obvious mutual interconnections become more reciprocal. Spreading failures are of particular interest. They usually originate in a single network segment and then spread to the rest of network often causing a global collapse. Two types of spreading failures are given focus, namely: epidemics and cascading failures. How to make backbone networks more resilient against spreading failures? How to tune the topology or additionally protect nodes or links in order to mitigate an effect of the potential failure? Those are the main questions addressed in this thesis. First, the epidemic phenomena are discussed. The subjects of epidemic modeling and identification of the most influential spreaders are addressed using a proposed Linear Time-Invariant (LTI) system approach. Throughout the years, LTI system theory has been used mostly to describe electrical circuits and networks. LTI is suitable to characterize the behavior of the system consisting of numerous interconnected components. The results presented in this thesis show that the same mathematical toolbox could be used for the complex network analysis. Then, cascading failures are discussed. Like any system which can be modeled using an interdependence graph with limited capacity of either nodes or edges, backbone networks are prone to cascades. Numerical simulations are used to model such failures. The resilience of European National Research and Education Networks (NREN) is assessed, weak points and critical areas of the network are identified and the suggestions for its modification are proposed.

Nous nous intéressons dans ce travail au problème de prévision de nouveaux liens dans des grands graphes de terrain. Nous explorons en particulier les approches topologiques dyadiques pour la prévision de liens. Différentes mesures de proximité topologique ont été étudiées dans la littérature pour prédire l’apparition de nouveaux liens. Des techniques d’apprentissage supervisé ont été aussi utilisées afin de combiner ces différentes mesures pour construire des modèles prédictifs. Le problème d’apprentissage supervisé est ici un problème difficile à cause notamment du fort déséquilibre de classes. Dans cette thèse, nous explorons différentes approches alternatives pour améliorer les performances des approches dyadiques pour la prévision de liens. Nous proposons d’abord, une approche originale de combinaison des prévisions fondée sur des techniques d’agrégation supervisée de listes triées (ou agrégation de préférences). Nous explorons aussi différentes approches pour améliorer les performances des approches supervisées pour la prévision de liens. Une première approche consiste à étendre l’ensemble des attributs décrivant un exemple (paires de noeuds) par des attributs calculés dans un réseau multiplexe qui englobe le réseau cible. Un deuxième axe consiste à évaluer l’apport destechniques de détection de communautés pour l’échantillonnage des exemples. Des expérimentations menées sur des réseaux réels extraits de la base bibliographique DBLP montrent l’intérêt des approaches proposées. / In this work, we are interested to tackle the problem of link prediction in complex networks. In particular, we explore topological dyadic approaches for link prediction. Different topological proximity measures have been studied in the scientific literature for finding the probability of appearance of new links in a complex network. Supervided learning methods have also been used to combine the predictions made or information provided by different topological measures. The create predictive models using various topological measures. The problem of supervised learning for link prediction is a difficult problem especially due to the presence of heavy class imbalance. In this thesis, we search different alternative approaches to improve the performance of different dyadic approaches for link prediction. We propose here, a new approach of link prediction based on supervised rank agregation that uses concepts from computational social choice theory. Our approach is founded on supervised techniques of aggregating sorted lists (or preference aggregation). We also explore different ways of improving supervised link prediction approaches. One approach is to extend the set of attributes describing an example (pair of nodes) by attributes calculated in a multiplex network that includes the target network. Multiplex networks have a layered structure, each layer having different kinds of links between same sets of nodes. The second way is to use community information for sampling of examples to deal with the problem of classe imabalance. Experiments conducted on real networks extracted from well known DBLP bibliographic database.

Les réseaux complexes, issus de relevés de terrain d’origines trèsvariées, en biologie, science de l’information ou sociologie,présentent une caractéristique remarquable dénommée structurecommunautaire. Des groupes, ou communautés, à l’intérieur duréseau, ont une cohésion interne forte et des liens entre eux plusfaibles. Sans connaissance a priori du nombre de communautés, ladifficulté réside dans la caractérisation d’un bon partitionnement encommunautés. La modularité est une mesure globale de qualité departitionnement très utilisée qui capture les contraintes de cohésioninterne forte et de liens externes faibles. Elle transforme le problèmede détection de communautés en problème d’optimisationNP-difficile. Elle souffre d’un défaut, la limite de résolution, qui tendà rendre indétectables les très petites communautés d’autant plusque le réseau est grand. L’algorithme le plus efficace pour optimiserla modularité, dit de Louvain, procède par fusion de communautés.Cette thèse s’attache à modifier cet algorithme pour qu’il réalisemajoritairement des fusions pertinentes, qui n’aggravent pas lalimite de résolution, en utilisant une condition de fusion. De plus, enl’associant à un algorithme mémétique, les partitions proposéessont très proches des partitions attendues pour des graphesgénérés par un modèle qui reproduit les caractéristiques desréseaux complexes. Enfin, cet algorithme mémétique réduitfortement l’inconsistance de solution, défaut de la modularité selonlequel deux partitions trouvées à partir d’un examen des noeudsdans un ordre aléatoire, pour le même graphe, peuvent êtrestructurellement très différentes, rendant leur interprétation délicate. / From various applications, in sociology or biology for instance,complex networks exhib the remarquable property of communitystructure. Groups, sometimes called communities, has a stronginternal cohesion and poor links between them. Whithout priorknowledge of the number of communities, the difficulty lies in thecharacterization of a good clustering. Modularity is an overallmeasure of clustering quality widely used to capture the doubleconstraint, internal and external, of well formed communities. Theproblem became a NP-hard optimization problem. The main weakof modularity is the resolution limit, which tends to makeundetectable very small communities especially as the network islarge. The algorithm of Louvain, one of the most efficient one tooptimize modularity, proceeds by merging communities. This thesisattempts to modify the algorithm so that it mainly produces relevantmerges that do not make worse the effects of resolution limit, usinga merge condition. In addition, by combining it with a memeticalgorithm, proposed clusterings are very close to the expected onesfor graphs generated by a model that reproduces the characteristicsof complex networks. Finally, the memetic algorithm greatly reducesthe inconsistency of solution, another weakness of modularity suchthat, for the same graph, two partitions found from an exploration ofnodes in a random order can be structurally very different, makingthem difficult to interpret.

Orientadores: José Antônio Brum, Marcus Aloizio Martinez de Aguiar / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Física Gleb Wataghin / Made available in DSpace on 2018-08-29T15:40:55Z (GMT). No. of bitstreams: 1 Reis_ElohimFonsecados_M.pdf: 2277988 bytes, checksum: 08f2c3b84a391217d575c0f425159fca (MD5) Previous issue date: 2015 / Resumo: Este trabalho é dividido em duas partes. Na primeira parte, uma rede de correlação é construída baseada em um modelo de Ising em diferentes temperaturas, crítica, subcrítica e supercrítica, usando um algorítimo de Metropolis Monte-Carlo com dinâmica de \textit{single-spin-flip}. Este modelo teórico é comparado com uma rede do cérebro construída a partir de correlações das séries temporais do sinal BOLD de fMRI de regiões do cérebro. Medidas de rede, como coeficiente de aglomeração, mínimo caminho médio e distribuição de grau são analisadas. As mesmas medidas de rede são calculadas para a rede obtida pelas correlações das séries temporais dos spins no modelo de Ising. Os resultados da rede cerebral são melhor explicados pelo modelo teórico na temperatura crítica, sugerindo aspectos de criticalidade na dinâmica cerebral. Na segunda parte, é estudada a dinâmica temporal da atividade de um população neural, ou seja, a atividade de células ganglionares da retina gravadas em uma matriz de multi-eletrodos. Vários estudos têm focado em descrever a atividade de redes neurais usando modelos de Ising com desordem, não dando atenção à estrutura dinâmica. Tratando o tempo como uma dimensão extra do sistema, a dinâmica temporal da atividade da população neural é modelada. O princípio de máxima entropia é usado para construir um modelo de Ising com interação entre pares das atividades de diferentes neurônios em tempos diferentes. O ajuste do modelo é feito com uma combinação de amostragem de Monte-Carlo e método do gradiente descendente. O sistema é caracterizado pelos parâmetros aprendidos, questões como balanço detalhado e reversibilidade temporal são analisadas e variáveis termodinâmicas, como o calor específico, podem ser calculadas para estudar aspectos de criticalidade / Abstract: This work is divided in two parts. In the first part, a correlation network is build based on an Ising model at different temperatures, critical, subcritical and supercritical, using a Metropolis Monte-Carlo algorithm with single-spin-flip dynamics. This theoretical model is compared with a brain network built from the correlations of BOLD fMRI temporal series of brain regions activity. Network measures, such as clustering coefficient, average shortest path length and degree distributions are analysed. The same network measures are calculated to the network obtained from the time series correlations of the spins in the Ising model. The results from the brain network are better explained by the theoretical model at the critical temperature, suggesting critical aspects in the brain dynamics. In the second part, the temporal dynamics of the activity of a neuron population, that is, the activity of retinal ganglion cells recorded in a multi-electrode array was studied. Many studies have focused on describing the activity of neural networks using disordered Ising models, with no regard to the dynamic nature. Treating time as an extra dimension of the system, the temporal dynamics of the activity of the neuron population is modeled. The maximum entropy principle approach is used to build an Ising model with pairwise interactions between the activities of different neurons at different times. Model fitting is performed by a combination of Metropolis Monte Carlo sampling with gradient descent methods. The system is characterized by the learned parameters, questions like detailed balance and time reversibility are analysed and thermodynamic variables, such as specific heat, can be calculated to study critical aspects / Mestrado / Física / Mestre em Física / 2013/25361-6 / FAPESP

The central theme of complex systems research is to understand the emergent macroscopic properties of a system from the interplay of its microscopic constituents. The emergence of macroscopic properties is often intimately related to the structure of the microscopic interactions. Here, we present an analytical approach for deriving necessary conditions that an interaction network has to obey in order to support a given type of macroscopic behaviour. The approach is based on a graphical notation, which allows rewriting Jacobi’s signature criterion in an interpretable form and which can be applied to many systems of symmetrically coupled units. The derived conditions pertain to structures on all scales, ranging from individual nodes to the interaction network as a whole. For the purpose of illustration, we consider the example of synchronization, specifically the (heterogeneous) Kuramoto model and an adaptive variant. The results complete and extend the previous analysis of Do et al. (2012 Phys. Rev. Lett. 108, 194102).

This thesis advances the theory of network specialization by characterizing the effect of network specialization on the eigenvectors of a network. We prove and provide explicit formulas for the eigenvectors of specialized graphs based on the eigenvectors of their parent graphs. The second portion of this thesis applies network specialization to learning problems. Our work focuses on training reservoir computers to mimic the Lorentz equations. We experiment with random graph, preferential attachment and small world topologies and demonstrate that the random removal of directed edges increases predictive capability of a reservoir topology. We then create a new network model by growing networks via targeted application of the specialization model. This is accomplished iteratively by selecting top preforming nodes within the reservoir computer and specializing them. Our generated topology out-preforms all other topologies on average.