Title :
A Hidden Trouble Theorem for Model Classing by BP Algorithm
Author :
Yang, Guowei ; Xing, Rong ; Wang, Shoujue
Author_Institution :
Qingdao Univ.
Abstract :
First, two lemmas for classing ability of BP neural network are given and proved. Then, a hidden trouble theorem for model classing by BP algorithm is proposed and proved, which clears misty cognition about classing ability of BP neural network and offers theoretic direction for improving the structure and algorithm of BP neural network
Keywords :
backpropagation; cognition; feedforward neural nets; pattern classification; backpropagation neural network; cognition; feedforward neural network; hidden trouble theorem; model classification; Artificial neural networks; Automation; Cognition; Electronic mail; Feedforward neural networks; Intelligent control; Laboratories; Neural networks; BP algorithm; BP neural network; feedforward neural network; model classing;
Conference_Titel :
Intelligent Control and Automation, 2006. WCICA 2006. The Sixth World Congress on
Conference_Location :
Dalian
Print_ISBN :
1-4244-0332-4
DOI :
10.1109/WCICA.2006.1712863
