京都大学 / 新制・課程博士 / 博士(情報学) / 甲第24039号 / 情博第795号 / 新制||情||134(附属図書館) / 京都大学大学院情報学研究科システム科学専攻 / (主査)教授 下平 英寿, 教授 田中 利幸, 教授 加納 学 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

The changing topology of a network is driven by the need to maintain or optimize network function. As this function is often related to moving quantities such as traffic, information, etc., efficiently through the network, the structure of the network and the dynamics on the network directly depend on the other. To model this interplay of network structure and dynamics we use the dynamics on the network, or the dynamical processes the network models, to influence the dynamics of the network structure, i.e., to determine where and when to modify the network structure. We model the dynamics on the network using Jackson network dynamics and the dynamics of the network structure using minimal specialization, a variant of the more general network growth model known as specialization. The resulting model, which we refer to as the integrated specialization model, coevolves both the structure and the dynamics of the network. We show this model produces networks with real-world properties, such as right-skewed degree distributions, sparsity, the small-world property, and non-trivial equitable partitions. Additionally, when compared to other growth models, the integrated specialization model creates networks with small diameter, minimizing distances across the network. Along with producing these structural features, this model also sequentially removes the network's largest bottlenecks. The result are networks that have both dynamic and structural features that allow quantities to more efficiently move through the network.

Sabe-se biologicamente que o nível de expressão dos genes está entre os fatores podem indicar o quanto estes estão em atividade em determinado momento. Avanços na tecnologia de microarray têm possibilitado medir os níveis de expressão de milhares de genes ao mesmo tempo. Esses dados podem ser medidos de maneira a formarem uma série temporal, que pode ser tratada estatisticamente para serem obtidas informações sobre as relações entre os genes. Já foram propostos vários modelos para tratar redes gênicas matematicamente. Esses modelos têm evoluído de forma a agregarem cada vez mais características das redes reais. Neste trabalho, será feita uma revisão de modelos discretos para redes de regulação gênica, primeiramente com as redes Booleanas, modelo determinístico, e depois as redes Booleanas probabilísticas e as redes genéticas probabilísticas, modelos que tratam o problema estocasticamente. Usando o último modelo citado, serão mostrados dois métodos para estimar o nível de predição entre os genes, coeficiente de determinação e informação mútua. Além de se estimar essas relações, foram desenvolvidas algumas técnicas para construir redes a partir de genes específicos, que são chamados sementes. Também serão apresentados dois desses métodos de crescimento de redes e, baseado neles, um terceiro método que foi desenvolvido neste trabalho. Foi criado um algoritmo que realiza o crescimento da rede mudando as sementes a cada iteração, agrupando estes genes em grupos com diferentes níveis de confiança, chamados camadas. O algoritmo também usa outros critérios para agregar novos genes à rede. Após a explanação desses métodos, será mostrado um software que, a partir de dados temporais de expressão gênica, estima as dependências entre os genes e executa o crescimento da rede em torno de genes que se deseje estudar. Também serão mostradas as melhorias feitas no programa. Ao final, serão apresentados alguns testes feitos com dados do Plasmodium falciparum, parasita causador da malária. / It\'s known that gene expression levels are among the factors that can show how genes are active in certain moment. Advances in microarray technology have given the possibility to measure expression levels of thousands of genes in a certain instant of time. These data constitute time series that we can treat statistically in order to get information genes relationships. Many models were proposed to treat gene networks mathematically. These models have evolved to aggregate more and more real networks features. In this work, it is made a brief review of discrete models of regulatory genetic networks, initially Boolean networks, a deterministic model, and then probabilistic Boolean networks and probabilistic genetic networks, models that treat the problem stochastically. Using the last model cited, two methods to estimate the prediction level among genes are shown, coefficient of determination and mutual information. Besides estimating these relations, some techniques have been developed to construct networks from specific genes, that are called seeds. It will be also shown two methods of network growth and, based on these, a third method that was developed during this work. An algorithm was created, such that it grows the network changing the seeds in each iteration, grouping these genes in groups with different level of confidence, called layers. The algorithm also uses other criteria to add new genes to the network. After studying these methods, it will be shown a software that, using time series gene expression data, estimates dependences among genes and runs the network growing process around chosen genes. It is also presented the improvements made in the program. Finally, some tests using data of Plasmodium falciparum, malaria parasite, are shown.

Sabe-se biologicamente que o nível de expressão dos genes está entre os fatores podem indicar o quanto estes estão em atividade em determinado momento. Avanços na tecnologia de microarray têm possibilitado medir os níveis de expressão de milhares de genes ao mesmo tempo. Esses dados podem ser medidos de maneira a formarem uma série temporal, que pode ser tratada estatisticamente para serem obtidas informações sobre as relações entre os genes. Já foram propostos vários modelos para tratar redes gênicas matematicamente. Esses modelos têm evoluído de forma a agregarem cada vez mais características das redes reais. Neste trabalho, será feita uma revisão de modelos discretos para redes de regulação gênica, primeiramente com as redes Booleanas, modelo determinístico, e depois as redes Booleanas probabilísticas e as redes genéticas probabilísticas, modelos que tratam o problema estocasticamente. Usando o último modelo citado, serão mostrados dois métodos para estimar o nível de predição entre os genes, coeficiente de determinação e informação mútua. Além de se estimar essas relações, foram desenvolvidas algumas técnicas para construir redes a partir de genes específicos, que são chamados sementes. Também serão apresentados dois desses métodos de crescimento de redes e, baseado neles, um terceiro método que foi desenvolvido neste trabalho. Foi criado um algoritmo que realiza o crescimento da rede mudando as sementes a cada iteração, agrupando estes genes em grupos com diferentes níveis de confiança, chamados camadas. O algoritmo também usa outros critérios para agregar novos genes à rede. Após a explanação desses métodos, será mostrado um software que, a partir de dados temporais de expressão gênica, estima as dependências entre os genes e executa o crescimento da rede em torno de genes que se deseje estudar. Também serão mostradas as melhorias feitas no programa. Ao final, serão apresentados alguns testes feitos com dados do Plasmodium falciparum, parasita causador da malária. / It\'s known that gene expression levels are among the factors that can show how genes are active in certain moment. Advances in microarray technology have given the possibility to measure expression levels of thousands of genes in a certain instant of time. These data constitute time series that we can treat statistically in order to get information genes relationships. Many models were proposed to treat gene networks mathematically. These models have evolved to aggregate more and more real networks features. In this work, it is made a brief review of discrete models of regulatory genetic networks, initially Boolean networks, a deterministic model, and then probabilistic Boolean networks and probabilistic genetic networks, models that treat the problem stochastically. Using the last model cited, two methods to estimate the prediction level among genes are shown, coefficient of determination and mutual information. Besides estimating these relations, some techniques have been developed to construct networks from specific genes, that are called seeds. It will be also shown two methods of network growth and, based on these, a third method that was developed during this work. An algorithm was created, such that it grows the network changing the seeds in each iteration, grouping these genes in groups with different level of confidence, called layers. The algorithm also uses other criteria to add new genes to the network. After studying these methods, it will be shown a software that, using time series gene expression data, estimates dependences among genes and runs the network growing process around chosen genes. It is also presented the improvements made in the program. Finally, some tests using data of Plasmodium falciparum, malaria parasite, are shown.

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.

No description available.

[...] Mes travaux portent sur les clusters structurant le réseau et l'innovation car 1) le cluster impacte collectivement plutôt qu’individuellement la sortie du réseau, 2) les couplages intra et inter-cluster représentent la structure même des clusters mais ils influencent différemment l'innovation ou la croissance du cluster, 3) un certain compromis reste à définir entre la structure dense et éparse des différents réseaux. Un cluster est de façon générale défini comme un groupe de choses similaires ou de personnes qui travaillent sur des sujets analogues. Selon le domaine auquel il s’applique, même si l’idée reste la même, la définition s’affine. En sciences des organisations, un cluster représente un regroupement d’entreprises et d’institutions qui interagissent entre-elles par le biais de contrats, d’opérations formelles ou informelles et de réunions occasionnelles afin de contribuer collectivement à un résultat innovant. [...] La thèse est structurée comme suit. Dans l'introduction générale, nous passons en revue la littérature des connaissances existantes qui sert de base pour le cadre conceptuel des documents. Nous définissons ensuite certains concepts utilisés dans les trois articles présentés tels que la structure de réseau complexe (utilisée dans le premier article), l'innovation et les liens de réseau (utilisés principalement dans le deuxième article), et la gestion des connaissances utilisées (dans le troisième article). Dans le premier article, nous discutons les différents mécanismes de formation de liens dictés par les réseaux dirigés permettant de distinguer la distribution des degrés. Dans le deuxième article, nous abordons l'impact de la dynamique de groupe sur l'innovation du groupe de projet OSS. Dans le troisième article, nous nous attachons à l'impact du transfert des connaissances à l'intérieur des groupes sur le transfert des connaissances entre les groupes. L'annexe A permettra de discuter la modélisation analytique de la croissance des réseaux sociaux en utilisant la projection de réseaux multicouches ; l'annexe B sera l’occasion de présenter statistiquement le lien entre les relations intragroupe et les relations intergroupe. / [...] However, there is a gap in the literature with regard to the analysis of cluster or group structure as an input and cluster or group innovation as an output, e.g. “impact of network cluster structure on cluster innovation and growth”, i.e. how intra- and inter-cluster coupling, structural holes and tie strength impact cluster innovation and growth; and how intra-cluster density affects inter-cluster coupling; that I address in my thesis.Therefore, I focus on the cluster (or group of individuals) rather than the individual to analyze both network structure and innovation, because 1) clusters represent collective impact on network output rather than individuals’ impact, 2) intra and inter cluster couplings both represent cluster structure but have different impacts on cluster innovation and growth, 3) trade-offs among dense and sparse network cluster structures are different from those associated with networks of individuals. [...] The thesis is structured as follows. In the general introduction, I review the literature of existing knowledge in the field, which serves as a basis for the conceptual framework for the papers. I then define certain concepts used in the papers, such as complex network structure used in the first paper, innovation and network ties mainly used in the second paper, and knowledge management used in the third paper. In the first paper I discuss directed networks’ different link formation mechanisms causing degree distribution distinction. In the second paper, I discuss the impact of group dynamics on OSS project group innovation. In the third paper, I discuss impact of knowledge transfer inside groups onto knowledge transfer between groups. In appendix A, I discuss analytical modeling of social network growth using multilayer network projection; and in appendix B, I discuss statistically how intragroup ties and intergroup ties are related.