Your browser doesn't support javascript.
loading
Fast adapting ensemble: a new algorithm for mining data streams with concept drift.
Ortíz Díaz, Agustín; del Campo-Ávila, José; Ramos-Jiménez, Gonzalo; Frías Blanco, Isvani; Caballero Mota, Yailé; Mustelier Hechavarría, Antonio; Morales-Bueno, Rafael.
Afiliação
  • Ortíz Díaz A; Department of Computer Science, University of Granma, 85100 Granma, Cuba.
  • del Campo-Ávila J; Department of Language and Computer Science, University of Málaga, Complejo Tecnológico, 29071 Málaga, Spain.
  • Ramos-Jiménez G; Department of Language and Computer Science, University of Málaga, Complejo Tecnológico, 29071 Málaga, Spain.
  • Frías Blanco I; Department of Computer Science, University of Granma, 85100 Granma, Cuba.
  • Caballero Mota Y; Department of Computer Science, University of Camagüey, 70100 Camagüey, Cuba.
  • Mustelier Hechavarría A; Department of Computer Science, University of Granma, 85100 Granma, Cuba.
  • Morales-Bueno R; Department of Language and Computer Science, University of Málaga, Complejo Tecnológico, 29071 Málaga, Spain.
ScientificWorldJournal ; 2015: 235810, 2015.
Article em En | MEDLINE | ID: mdl-25879051
The treatment of large data streams in the presence of concept drifts is one of the main challenges in the field of data mining, particularly when the algorithms have to deal with concepts that disappear and then reappear. This paper presents a new algorithm, called Fast Adapting Ensemble (FAE), which adapts very quickly to both abrupt and gradual concept drifts, and has been specifically designed to deal with recurring concepts. FAE processes the learning examples in blocks of the same size, but it does not have to wait for the batch to be complete in order to adapt its base classification mechanism. FAE incorporates a drift detector to improve the handling of abrupt concept drifts and stores a set of inactive classifiers that represent old concepts, which are activated very quickly when these concepts reappear. We compare our new algorithm with various well-known learning algorithms, taking into account, common benchmark datasets. The experiments show promising results from the proposed algorithm (regarding accuracy and runtime), handling different types of concept drifts.

Texto completo: 1 Coleções: 01-internacional Base de dados: MEDLINE Tipo de estudo: Prognostic_studies Idioma: En Revista: ScientificWorldJournal Assunto da revista: MEDICINA Ano de publicação: 2015 Tipo de documento: Article País de afiliação: Cuba País de publicação: Estados Unidos

Texto completo: 1 Coleções: 01-internacional Base de dados: MEDLINE Tipo de estudo: Prognostic_studies Idioma: En Revista: ScientificWorldJournal Assunto da revista: MEDICINA Ano de publicação: 2015 Tipo de documento: Article País de afiliação: Cuba País de publicação: Estados Unidos