Title :
On the design of constructive training algorithms for multilayer feedforward
Author :
Hernández-Espinosa, Carlos ; Fernández-Redondo, Mercedes
Author_Institution :
Departamento de Ingenieria y Ciencia de los Computadores, Univ. Jaume I, Castellon, Spain
fDate :
6/24/1905 12:00:00 AM
Abstract :
The usual way to determine the number of hidden units in multilayer feedforward networks is by a trial and error procedure. An alternative is to use constructive algorithms which change the neural network structure during training. In this paper we propose a new constructive algorithm, explore several design alternatives and conclude that this new constructive algorithm is a better alternative than the procedure of trial and error for designing a neural network
Keywords :
backpropagation; convergence; feedforward neural nets; learning (artificial intelligence); backpropagation; constructive algorithms; convergence; feedforward neural network; hidden layer; learning; multilayer neural network; Algorithm design and analysis; Bibliographies; Computer networks; Convergence; Electrochemical machining; Electronic mail; Multi-layer neural network; Neural networks; Neurons; Nonhomogeneous media;
Conference_Titel :
Neural Networks, 2002. IJCNN '02. Proceedings of the 2002 International Joint Conference on
Conference_Location :
Honolulu, HI
Print_ISBN :
0-7803-7278-6
DOI :
10.1109/IJCNN.2002.1005592
