Title :
A simple adaptive back-propagation algorithm for multilayered feedforward perceptrons
Author :
Cerqueira, Jes J. F.
Author_Institution :
Departamento de Engenharia Eletrica, Univ. Fed. da Bahia - UFBA, Salvador, Brazil
Abstract :
The paper presents a simple adaptive backpropagation algorithm based on an upper bound for the learning rate. This upper bound maintains the algorithm convergence and only external modifications need to be made in the backpropagation computational routines existent.
Keywords :
backpropagation; feedforward neural nets; multilayer perceptrons; algorithm convergence; backpropagation computational routines; learning rate; multilayered feedforward perceptrons; simple adaptive backpropagation algorithm; upper bound; Computer hacking; Convergence; Gradient methods; Helium; Multilayer perceptrons; Nonlinear systems; Optimal control; Optimization methods; Taylor series; Upper bound;
Conference_Titel :
Systems, Man and Cybernetics, 2002 IEEE International Conference on
Print_ISBN :
0-7803-7437-1
DOI :
10.1109/ICSMC.2002.1176108
