%0 Journal Article
%T A Geometrical Strategy of Constructive Initial Neural Networks<br>神经网络设计的特征空间序贯划分算法
%A SUN Gong-Xing DAI Gui-Liang
%A <br>孙功星
%A 戴贵亮
%J 计算机科学
%D 2003
%I 
%X A geometrical strategy for constructive neural networks is proposed in the paper. Firstly it can acquire quickly initial input weight parameters and topolgy by sequentially partioning feature space with the presented geometrical method. Secondly with SVD,its initial output weights are obtained very quickly. Finally these weights are retuned with BP algorithms. Its distinctive features are that it can construct quickly an initial neural networks using geometrical method other than backpropagation algorithms so that overtraining and undertraining are avoided automatically,and experimentally it performs better on two-spiral classification than Cascade-Correlation Algorithm.
%K Geometrical stragegy
%K Overtraining
%K Sigular value decomposition<br>神经网络
%K 特征空间
%K 序贯划分算法
%K 误差反转学习算法
%K 奇异值分解方法
%U http://www.alljournals.cn/get_abstract_url.aspx?pcid=5B3AB970F71A803DEACDC0559115BFCF0A068CD97DD29835&cid=8240383F08CE46C8B05036380D75B607&jid=64A12D73428C8B8DBFB978D04DFEB3C1&aid=52E681F9571457EA&yid=D43C4A19B2EE3C0A&vid=340AC2BF8E7AB4FD&iid=708DD6B15D2464E8&sid=933658645952ED9F&eid=42425781F0B1C26E&journal_id=1002-137X&journal_name=计算机科学&referenced_num=0&reference_num=9