Exportar registro bibliográfico

Exploration of neural network techniques to detect community structure in complex networks (2021)

  • Authors:
  • USP affiliated author: GOMAZAKO, NATHALIA SATIE - IFSC
  • School: IFSC
  • Subjects: REDES COMPLEXAS; REDES NEURAIS
  • Language: Inglês
  • Abstract: Recently, complex networks have been successfully used for modeling real-world systems as sets of elements (the nodes) with pairwise connections (the links). The topology of this network frequently contains important information about the system. One topological feature that is common in complex networks is its community structure. A community is a set of nodes that is more strongly connected among themselves than with nodes outside the community. Detecting communities by trying different combinations of nodes is not viable for large networks, and thus heuristic methods are generally used. Finding a division of the nodes of a network in communities corresponds to assigning to each node a label, the number of the community to which it belongs. This means that community detection can be seen as a classification problem. Neural networks are generally a useful tool for classification. This suggests the use of neural networks for the detection of communities in complex networks, which is the purpose of this work. We train and test a neural network with two different models of complex networks with community structure and evaluate the accuracy of the community structure detected. We also compare the results of the neural network with some well established community detection algorithms. We find that the neural network has good results for the simple models used, comparable or even superior to that of well established algorithms, but the accuracy decreases when a more flexible community generation model is used. Also, the accuracy is strongly affected by the model used for training (that is, by the kind of community structure present in the training set), which is expected, but implies that careful attention must be given to the training set
  • Imprenta:

  • Download do texto completo

    Tipo Nome Link
    Versão PublicadaNathalia Satie Gomazako f...Direct link
    How to cite
    A citação é gerada automaticamente e pode não estar totalmente de acordo com as normas

    • ABNT

      GOMAZAKO, Nathalia Satie. Exploration of neural network techniques to detect community structure in complex networks. 2021. Trabalho de Conclusão de Curso (Graduação) – Instituto de Física de São Carlos, Universidade de São Paulo, São Carlos, 2021. Disponível em: https://repositorio.usp.br/directbitstream/143da3f3-8243-488e-a142-24c298274a19/Nathalia%20Satie%20Gomazako%20final.pdf. Acesso em: 04 dez. 2022.
    • APA

      Gomazako, N. S. (2021). Exploration of neural network techniques to detect community structure in complex networks (Trabalho de Conclusão de Curso (Graduação). Instituto de Física de São Carlos, Universidade de São Paulo, São Carlos. Recuperado de https://repositorio.usp.br/directbitstream/143da3f3-8243-488e-a142-24c298274a19/Nathalia%20Satie%20Gomazako%20final.pdf
    • NLM

      Gomazako NS. Exploration of neural network techniques to detect community structure in complex networks [Internet]. 2021 ;[citado 2022 dez. 04 ] Available from: https://repositorio.usp.br/directbitstream/143da3f3-8243-488e-a142-24c298274a19/Nathalia%20Satie%20Gomazako%20final.pdf
    • Vancouver

      Gomazako NS. Exploration of neural network techniques to detect community structure in complex networks [Internet]. 2021 ;[citado 2022 dez. 04 ] Available from: https://repositorio.usp.br/directbitstream/143da3f3-8243-488e-a142-24c298274a19/Nathalia%20Satie%20Gomazako%20final.pdf

    Últimas obras dos mesmos autores vinculados com a USP cadastradas na BDPI:

    Digital Library of Academic Works of Universidade de São Paulo     2012 - 2022