[1]赵勇,冯纯伯.时变最优化信息处理技术及其应用(Ⅱ)——前馈神经网络学习算法[J].东南大学学报(自然科学版),1999,29(4):98-102.[doi:10.3969/j.issn.1001-0505.1999.04.021]
　Zhao Yong,Feng Chunbo.Time-Dependent Optimization for Information Processing and Its Applications(Ⅱ) On-Line Neural Network Learning[J].Journal of Southeast University (Natural Science Edition),1999,29(4):98-102.[doi:10.3969/j.issn.1001-0505.1999.04.021]
点击复制

时变最优化信息处理技术及其应用(Ⅱ)——前馈神经网络学习算法()
分享到：

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
29
期数:
1999年第4期
页码:
98-102
栏目:
自动化
出版日期:
1999-07-20

文章信息/Info

Title:
Time-Dependent Optimization for Information Processing and Its Applications(Ⅱ) On-Line Neural Network Learning
作者:
赵勇1;  冯纯伯2
1 杭州大学电子工程系,杭州 310028; 2 东南大学自动化研究所,南京 210096
Author(s):
Zhao Yong1;  Feng Chunbo2
1 Department of Electronic Engineering, Hangzhou University, Hangzhou 310028
2 Research Institute of Automation, Southeast University, Nanjing 210096
关键词:
时变最优化;  信息处理;  神经网络
Keywords:
time-varying optimization;  information processing;  neural networks
分类号:
TP14
DOI:
10.3969/j.issn.1001-0505.1999.04.021
摘要:
多层前馈神经网络应用成功的关键之一在于寻找一种有效的学习方法. 尤其是在线应用,其学习速度和效率显得极为重要. 而目前应用最广泛的BP算法却存在收敛慢和振荡等缺点. 应用时变最优化技术,可以得到以指数速率收敛的前馈神经网络在线学习算法,理论分析和仿真表明该方法具有较快的收敛速度,并能迅速学习或适应新的输入输出模式.
Abstract:
The key to successful applications of multi-layer feedforward neural networks is to find out an efficient learning algorithm. This is extremely important especially in on-line applications. The disadvantages of the most widely used BP algorithm are slow convergence and oscillation. With time-dependent optimization, on-line feedforward neural network learning algorithm with exponential convergence rate is derived. Both theoretical analysis and simulations show that this algorithm has a fast convergence rate and enable the network to learn and adapt to new input-output patterns quickly.

参考文献/References:

[1] Rumlhart D E,Hinton G E,William R J.Parallel distribution processing,chapter 8:Learning internal representations.MIT Press,1996
[2] Becker S.Improving the convergence of back-propagation learning with second order methods.In:Proc 1988 Connectionist Models Summer School.Camegie-Mellon University,1988
[3] Yel S J,Stack H.A fast learning algorithm for multilayer neural network based on projection methods.Neural Network Theory and Applications,Academic Press,1991
[4] Zhao Y,Lu X W.Training neural networks using time-varying optimization.In:Proc of Int Conf on Neural Networks and Signal Processing’93,Guangzhou,1993

相似文献/References:

[1]孙志挥.DSK-1系统绘图语言设计及其实现[J].东南大学学报(自然科学版),1980,10(3):23.[doi:10.3969/j.issn.1001-0505.1980.03.003]
　Sun Zhi-hui.Design and Implementation of Drawing Language of DSK-1 System[J].Journal of Southeast University (Natural Science Edition),1980,10(4):23.[doi:10.3969/j.issn.1001-0505.1980.03.003]
[2]赵勇,冯纯伯.时变最优化信息处理技术及其应用(Ⅰ)——基本思想与系统辨识[J].东南大学学报(自然科学版),1999,29(4):92.[doi:10.3969/j.issn.1001-0505.1999.04.020]
　Zhao Yong,Feng Chunbo.Time-Dependent Optimization for Information Processing and Its Applications(Ⅰ)[J].Journal of Southeast University (Natural Science Edition),1999,29(4):92.[doi:10.3969/j.issn.1001-0505.1999.04.020]

备注/Memo

备注/Memo:
基金项目:国家自然科学基金资助项目(A-6319811044).
第一作者:男, 1963年生, 博士, 副教授.
更新日期/Last Update: 1999-07-20