Título Evolutionary combining of basis function neural networks for classification
Autores Hervás C., MARTÍNEZ ESTUDILLO, FRANCISCO JOSÉ, CARBONERO RUZ, MARIANO, Romero C., Fernández J.C., CARBONERO RUZ, MARIANO, MARTÍNEZ ESTUDILLO, FRANCISCO JOSÉ
Publicación externa No
Medio Lect. Notes Comput. Sci.
Alcance Conference Paper
Naturaleza Científica
Cuartil JCR 4
Cuartil SJR 2
Impacto SJR 0.29300
Ámbito Internacional
Web https://www.scopus.com/inward/record.uri?eid=2-s2.0-38149050564&partnerID=40&md5=4a955fd79e10c91bc2b9ca7ef723ebcb
Fecha de publicacion 01/01/2007
Scopus Id 2-s2.0-38149050564
Abstract The paper describes a methodology for constructing a possible combination of different basis functions (sigmoidal and product) for the hidden layer of a feed forward neural network, where the architecture, weights and node typology are learned based on evolutionary programming. This methodology is tested using simulated Gaussian data set classification problems with different linear correlations between input variables and different variances. It was found that combined basis functions are the more accurate for classification than pure sigmoidal or product-unit models. Combined basis functions present competitive results which are obtained using linear discriminant analysis, the best classification methodology for Gaussian data sets. © Springer-Verlag Berlin Heidelberg 2007.
Palabras clave Computer simulation; Correlation methods; Discriminant analysis; Evolutionary algorithms; Gaussian distribution; Combined basis functions; Gaussian data set classification; Gaussian data sets; Radial
Miembros de la Universidad Loyola