Rank-based self-training for graph convolutional networks

Registro completo de metadados
MetadadosDescriçãoIdioma
Autor(es): dc.contributorUniversidade Estadual Paulista (Unesp)-
Autor(es): dc.contributorTemple University-
Autor(es): dc.creatorPedronette, Daniel Carlos Guimarães [UNESP]-
Autor(es): dc.creatorLatecki, Longin Jan-
Data de aceite: dc.date.accessioned2022-02-22T00:48:39Z-
Data de disponibilização: dc.date.available2022-02-22T00:48:39Z-
Data de envio: dc.date.issued2021-06-25-
Data de envio: dc.date.issued2021-06-25-
Data de envio: dc.date.issued2021-03-01-
Fonte completa do material: dc.identifierhttp://dx.doi.org/10.1016/j.ipm.2020.102443-
Fonte completa do material: dc.identifierhttp://hdl.handle.net/11449/206925-
Fonte: dc.identifier.urihttp://educapes.capes.gov.br/handle/11449/206925-
Descrição: dc.descriptionGraph Convolutional Networks (GCNs) have been established as a fundamental approach for representation learning on graphs, based on convolution operations on non-Euclidean domain, defined by graph-structured data. GCNs and variants have achieved state-of-the-art results on classification tasks, especially in semi-supervised learning scenarios. A central challenge in semi-supervised classification consists in how to exploit the maximum of useful information encoded in the unlabeled data. In this paper, we address this issue through a novel self-training approach for improving the accuracy of GCNs on semi-supervised classification tasks. A margin score is used through a rank-based model to identify the most confident sample predictions. Such predictions are exploited as an expanded labeled set in a second-stage training step. Our model is suitable for different GCN models. Moreover, we also propose a rank aggregation of labeled sets obtained by different GCN models. The experimental evaluation considers four GCN variations and traditional benchmarks extensively used in the literature. Significant accuracy gains were achieved for all evaluated models, reaching results comparable or superior to the state-of-the-art. The best results were achieved for rank aggregation self-training on combinations of the four GCN models.-
Descrição: dc.descriptionMicrosoft Research-
Descrição: dc.descriptionFundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)-
Descrição: dc.descriptionConselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)-
Descrição: dc.descriptionNational Science Foundation-
Descrição: dc.descriptionDepartment of Statistics Applied Mathematics and Computing (DEMAC) São Paulo State University (UNESP)-
Descrição: dc.descriptionDepartment of Computer and Information Sciences Temple University-
Descrição: dc.descriptionDepartment of Statistics Applied Mathematics and Computing (DEMAC) São Paulo State University (UNESP)-
Descrição: dc.descriptionFAPESP: #2017/25908-6-
Descrição: dc.descriptionFAPESP: #2018/15597-6-
Descrição: dc.descriptionCNPq: #308194/2017-9-
Descrição: dc.descriptionNational Science Foundation: IIS-1814745-
Idioma: dc.languageen-
Relação: dc.relationInformation Processing and Management-
???dc.source???: dc.sourceScopus-
Palavras-chave: dc.subjectGraph convolutional networks-
Palavras-chave: dc.subjectRank model-
Palavras-chave: dc.subjectSelf-training-
Palavras-chave: dc.subjectSemi-supervised learning-
Título: dc.titleRank-based self-training for graph convolutional networks-
Tipo de arquivo: dc.typelivro digital-
Aparece nas coleções:Repositório Institucional - Unesp

Não existem arquivos associados a este item.