Discriminant eigenfaces: A new ranking method for principal components analysis

Nenhuma Miniatura disponível
Citações na Scopus
1
Tipo de produção
Artigo de evento
Data
2008-10-26
Autores
Carlos E. Thomaz
Giraldi G. A.
Orientador
Periódico
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Título da Revista
ISSN da Revista
Título de Volume
Citação
THOMAZ, C. E.; GIRALDI, G. A. Discriminant eigenfaces: A new ranking method for principal components analysis. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), v. 5249 LNAI, p. 43-52, Oct, 2008.
Texto completo (DOI)
Palavras-chave
Resumo
Principal Component Analysis (PCA) is one of the most successful approaches to the problem of creating a low dimensional data representation and interpretation. However, since PCA explains the covariance structure of all the data, the first principal components with the largest eigenvalues do not necessarily represent important discriminant directions to separate sample groups. In this work, we investigate a new ranking method for the principal components. Instead of sorting the principal components in decreasing order of the corresponding eigenvalues, we propose the idea of using the discriminant weights given by separating hyperplanes to select among the principal components the most discriminant ones. Our experimental results have shown that the principal components selected by the separating hyperplanes are quite useful for understanding the differences between sample groups in face image analysis, allowing robust reconstruction and interpretation of the data as well as higher recognition rates using less linear features. © 2008 Springer Berlin Heidelberg.

Coleções