DocumentCode :
1752828
Title :
A Fast Learning Strategy for Multilayer Feedforward Neural Networks
Author :
Chen, Huawei ; Zhong, Hualan ; Yuan, Haiying ; Jin, Fan
Author_Institution :
Sch. of Inf. Sci. & Technol., Southwest Jiaotong Univ., Chengdu
Volume :
1
fYear :
0
fDate :
0-0 0
Firstpage :
3019
Lastpage :
3023
Abstract :
This paper proposes a new training algorithm called bi-phases weights´ adjusting (BPWA) for feedforward neural networks. Unlike BP learning algorithm, BPWA can adjust the weights during both forward phase and backward phase. The algorithm computes the minimum norm square solution as the weights between the hidden layer and output layer in the forward pass, while the backward pass, on the other hand, adjusts other weights in the network according to error gradient descent method. The experimental results based on function approximation and classification tasks show that new algorithm is able to achieve faster converging speed with good generalization performance when compared with the BP and Levenberg-Marquardt BP algorithm
Keywords :
feedforward neural nets; function approximation; gradient methods; learning (artificial intelligence); least squares approximations; multilayer perceptrons; pattern classification; biphases weights adjusting; classification tasks; error gradient descent; function approximation; learning strategy; minimum norm least-squares solution; multilayer feedforward neural networks; training algorithm; Approximation algorithms; Automation; Computer networks; Educational institutions; Electronic mail; Feedforward neural networks; Information science; Multi-layer neural network; Neural networks; Paper technology; bi-phases weights´ adjusting; feedforward neural network; minimum norm least-squares solution;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Intelligent Control and Automation, 2006. WCICA 2006. The Sixth World Congress on
Conference_Location :
Dalian
Print_ISBN :
1-4244-0332-4
Type :
conf
DOI :
10.1109/WCICA.2006.1712920
Filename :
1712920
Link To Document :
بازگشت