首页 | 本学科首页   官方微博 | 高级检索  
   检索      

GAL神经网络训练算法的改进
引用本文:晁勇,丁汉哲,刘福伟.GAL神经网络训练算法的改进[J].装甲兵工程学院学报,1999(3).
作者姓名:晁勇  丁汉哲  刘福伟
作者单位:装甲兵工程学院运用与维修工程教研室 北京100072
摘    要:对一种新型的神经网络——GAL(Grow and Learn)神经网络给出一种改进训练算法,与未改进的网络在训练时间、识别时间及识别准确率等方面进行比较,证明该网络结构简单、隐层节点易调整、运算量小,且识别准确率并未降低。

关 键 词:神经网络  模式识别  信号检测

Development of a Training Algorithm for GAL Neural Network
Chao Yong Ding Hanzhe Liu Fuwei.Development of a Training Algorithm for GAL Neural Network[J].Journal of Armored Force Engineering Institute,1999(3).
Authors:Chao Yong Ding Hanzhe Liu Fuwei
Institution:Chao Yong Ding Hanzhe Liu Fuwei
Abstract:The paper describes a new training algorithm for GAL (Grow and Learn) neural network, with an easy-changeable structure. With a comparison to the former training algorithm, it shows that the GAL neural network has the characteristics of simple structure, fast learning and the same accuracy rate.
Keywords:neural networks pattern recognition signal detection  
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号