An Empirical Methodology to Analyze the Behavior of Bagging

• Tipo: Artigo em Livro de Atas de Conferência Internacional
• Data: 2014

Autores

Pinto, F (Autor) Outra; Carlos Soares (Autor) FEUP; João Mendes-Moreira (Autor) FEUP

Ata de Conferência Internacional

Título: ADVANCED DATA MINING AND APPLICATIONS, ADMA 2014

Páginas: 199-212

10th International Conference on Advanced Data Mining and Applications (ADMA)

Guilin, PEOPLES R CHINA, DEC 19-21, 2014

Indexação

ISI Web of Knowledge - 1 Citação

Scopus - 2 Citações

Classificação Científica

CORDIS: Ciências Físicas > Ciência de computadores > Cibernética > Inteligência artificial

FOS: Ciências exactas e naturais > Ciências da computação e da informação

Outras Informações

ID Authenticus: P-00A-1F1

DOI: 10.1007/978-3-319-14717-8_16

Abstract (EN): In this paper we propose and apply a methodology to study the relationship between the performance of bagging and the characteristics of the bootstrap samples. The methodology consists of 1) an extensive set of experiments to estimate the empirical distribution of performance of the population of all possible ensembles that can be created with those bootstraps and 2) a metalearning approach to analyze that distribution based on characteristics of the bootstrap samples and their relationship with the complete training set. Given the large size of the population of all ensembles, we empirically show that it is possible to apply the methodology to a sample. We applied the methodology to 53 classification datasets for ensembles of 20 and 100 models. Our results show that diversity is crucial for an important bootstrap and we show evidence of a metric that can measure diversity without any learning process involved. We also found evidence that the best bootstraps have a predictive power very similar to the one presented by the training set using naive models.

Idioma: Inglês

Tipo (Avaliação Docente): Científica

Nº de páginas: 14