Title :
On the composite squared-error algorithm for neural networks
Author :
Netto, Sergio L. ; De Campos, Marcello L R
Author_Institution :
COPPE, Univ. Fed. do Rio de Janeiro, Brazil
Abstract :
The composite squared-error (CSE) algorithm results from the combination of the backpropagation algorithm with the pseudo-linear algorithm of Scalero and Tepedelenlioglu (1992). It is observed that the CSE algorithm is able to avoid suboptimal solutions and associated saddle points, thus achieving lower values of the error function than the pseudo-linear algorithm, in fewer iterations than the backpropagation algorithm. In this paper, we investigate the implementation of the CSE algorithm with the concepts of momentum gain, time-varying learning rate, and time-varying combining factor. It is verified that these features can improve the overall properties of the CSE convergence process
Keywords :
convergence; learning (artificial intelligence); neural nets; backpropagation algorithm; composite squared-error algorithm; convergence process; error function; momentum gain; neural networks; pseudo-linear algorithm; time-varying combining factor; time-varying learning rate; Backpropagation algorithms; Convergence; Error correction; Feedforward neural networks; Neural networks; Neurons; Signal processing; Signal processing algorithms; Steady-state;
Conference_Titel :
Circuits and Systems, 2000. Proceedings. ISCAS 2000 Geneva. The 2000 IEEE International Symposium on
Conference_Location :
Geneva
Print_ISBN :
0-7803-5482-6
DOI :
10.1109/ISCAS.2000.856018
