info presentación miembros líneas investigación publicaciones investigación tesis docencia

[Articulos Revista] [Articulos Congreso] [Reports] [Libros] [Capitulos Libro] [Todo] [Resumen]


Autores
Categoria WoS
Area
  • Artificial Neural Networks
  • Genetic Algorithms
Titulo: Automatic Selection of the Training Set with Genetic Algorithm for Training Artificial Neural Networks
Tipo: congreso internacional
Congreso: Genetic and Evolutionary Computation Conference (GECCO 2000)
Fecha: 10-12/7/2000
Lugar celebracion: Las Vegas, Nevada (EEUU)
Paginas: 64-66
ISBN: 1-55860-708-0
Libro: Proceedings of GECCO 2000
Editorial: Morgan Kaufmann Publishers

Abstract:

The processing of the data set used both in the trainig stage and in the test of an ANN is a simple filtering made to eliminate atypical values and to soothe the high frequency noise wich may slow down the ANN's learning process. This paper introduces a method for the selection of an optimal training set taken from time series that complements the already existing methods for the discrimination of input variables.

SABIA
    .: SABIA :.  Sistemas Adaptativos y Bioinspirados en Inteligencia Artificial