Consistent Model Selection for the Degree Corrected Stochastic Blockmodel

Autor(es)

Cerqueira, Andressa ; Gallo, Sandro ; Leonardi, Florencia ; Vera, Cristel

• Tipo de documento: Artigo Científico
• Fonte: ALEA-LATIN AMERICAN JOURNAL OF PROBABILITY AND MATHEMATICAL STATISTICS; v. 21, p. 26-pg., 2024-01-01.

Resumo

The Degree Corrected Stochastic Block Model (DCSBM) is a probabilistic model for random networks with community structure in which vertices of the same community are allowed to have distinct degree distributions. On the modeling side, this property makes the DCSBM more suitable for real-life complex networks. On the statistical side, it is more challenging due to a large number of parameters. In this paper, we prove that the penalized marginal likelihood estimator, when assuming prior distributions for the parameters, is strongly consistent for estimating the number of communities. We consider dense or semi-sparse random networks, and our estimator is unbounded, in the sense that the number of communities k considered can be as big as n, the number of nodes in the network. (AU)

• Processo FAPESP: 13/07699-0 - Centro de Pesquisa, Inovação e Difusão em Neuromatemática - NeuroMat
• Beneficiário: Oswaldo Baffa Filho
• Modalidade de apoio: Auxílio à Pesquisa - Centros de Pesquisa, Inovação e Difusão - CEPIDs
• Processo FAPESP: 17/10555-0 - Modelagem estocástica de sistemas interagentes
• Beneficiário: Fabio Prates Machado
• Modalidade de apoio: Auxílio à Pesquisa - Temático
• Processo FAPESP: 19/23439-4 - Cadeias de longo alcance e passeios aleatórios em grafos
• Beneficiário: Alexsandro Giacomo Grimbert Gallo
• Modalidade de apoio: Auxílio à Pesquisa - Regular
• Processo FAPESP: 23/05857-9 - Análise Estatística de Redes Complexas
• Beneficiário: Andressa Cerqueira
• Modalidade de apoio: Auxílio à Pesquisa - Regular