Título: Uncovering Overlapping Structures via Stochastic Competitive Learning
Autores: Silva, Thiago C.; Zhao, Liang
Resumo: Competitive learning is an important approach in Machine Learning. In this paper, we present a method for determining overlapping structures or vertices in the network using a stochastic competitive model, where several particles walk in the network and compete with each other to occupy as many nodes as possible, while attempting to reject intruder particles. The proposed measure for detecting overlapping structures is built from the rich information that is inherently embedded within the model description. Therefore, no extra processing is necessary to detect the overlapping structures in the data. Computer simulations reveal that the proposed overlapping index works well in real-world data sets.
Palavras-chave: Stochastic competitive learning; overlapping vertices; unsupervised learning
Código DOI: 10.21528/CBIC2011-23.1
Artigo em pdf: st_23.1.pdf
Arquivo BibTex: st_23.1.bib