An effective way to generate neural network structures for function approximation

One still open question in the area of research of multi-layer feedforward neural networks is concerning the number of neurons in its hidden layer(s). Especially in real life applications, this problem is often solved by heuristic methods. In this work an effective way to dynamically determine the n...

Descripción completa

Detalles Bibliográficos
Autor: Bastian, A.
Tipo de recurso: artículo
Fecha de publicación:1994
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2099/2445
Acceso en línea:https://hdl.handle.net/2099/2445
Access Level:acceso abierto
Palabra clave:Neural networks
Xarxes neuronals (Informàtica)
Classificació AMS::68 Computer science::68T Artificial intelligence
Descripción
Sumario:One still open question in the area of research of multi-layer feedforward neural networks is concerning the number of neurons in its hidden layer(s). Especially in real life applications, this problem is often solved by heuristic methods. In this work an effective way to dynamically determine the number of hidden units in a three-layer feedforward neural network for function approximation is proposed.