[1]乔俊飞,张 颖.一种多层前馈神经网络的快速修剪算法[J].智能系统学报,2008,3(02):173-176.
 QIAO Jun-fei,ZHANG Ying.Fast unit pruning algorithm for multilayer feedforward network design[J].CAAI Transactions on Intelligent Systems,2008,3(02):173-176.
点击复制

一种多层前馈神经网络的快速修剪算法(/HTML)
分享到:

《智能系统学报》[ISSN:1673-4785/CN:23-1538/TP]

卷:
第3卷
期数:
2008年02期
页码:
173-176
栏目:
出版日期:
2008-04-25

文章信息/Info

Title:
Fast unit pruning algorithm for multilayer feedforward network design
文章编号:
1673-4785(2008)02-0173-04
作者:
乔俊飞张  颖
北京工业大学电子信息与控制工程学院,北京100022
Author(s):
QIAO Jun-fei ZHANG Ying
College of Electronic and Control Engineering, Beijing University of Technology , Beijing 100022,China
关键词:
最优脑外科算法神经网络修剪算法自组织设计算法
Keywords:
optimal brain surgeon neural network pruning selforganization desig n
分类号:
TP183
文献标志码:
A
摘要:
针对目前神经网络在应用中难于确定隐层神经元数的问题,提出了一种神经网络结构的快速修剪算法.该算法在最优脑外科算法(OBS)的基础上,通过直接剔除冗余的隐层神经元实现神经网络结构自组织设计.实验结果表明,快速修剪算法与常规的最优脑外科算法相比,具有更简单的网络结构和更快的学习速度.
Abstract:
For it is difficult to determine the numbers of hidden neurons in the application of neural networks, a fast unit pruning algorithm for the structure of neural network was presented in the paper. The algorithm which based on optim al brain surgeon(OBS)eliminated the unneeded hidden neurons directly, in which way carried out the selforganization design on the structure of neural networ k s. The results of comparative studies with OBS showed that the fast unit pruning algorithm could reduce both neural network complexity and training time.

参考文献/References:

[1]魏海坤,徐嗣鑫,宋文忠.神经网络的泛化理论和泛化方法[J].自动化学报, 2001,27( 6):806815.
 WEI Haikun, XU Sixin,SONG Wenzhong. Generalization theory and generalization met hods for neural networks [J]. Acta Automatica Sinica, 2001, 27(6): 806815.
[2]何述东.多层前向神经网络结构的研究进展[J].控制理论与应用,1998, 15(3 ):313319.
 HE Shudong. Survey of architecture for multilayer feedfor ward neural networks [J]. Control Theory and Applications, 1998,15(3):313319.
[3]MOODY J. Prediction risk and architecture selection for neural net works [C]// Statistics to Neural Networks: Theory and Pattern Recognition Applications, NATO ASI Series F. New York, 1994.
[4]FAHLMAN S E , LEBIERE C. The cascadecorrelation learning architecture [C]// Advances in Neural Information Processing Systems. San Mateo,USA,1990 .
[5]HASSIBI B. STORK D,WOLFF G. Optimal brain surgeon and general network prun ing[C]// IEEE International Conferenceon on Neural Networks. Perth,Australia,1993. 
[6]杨钟瑾,史忠科.快速自顶向下优化神经网络结构的方法[J].系统仿真学报,2 005,17(1):162165.
YANG Zhongjin, SHI Zhongke. Fast approach for optimal brain surgeon[J]. Journ a l of System Simulation,2005,17(1):162165.
[7]李 倩,王永县,朱友芹. 人工神经网络混合剪枝算法[J]. 清华大学学报, 2005,45(6):831834.
LI Qian, WANG Yongxian,ZHU Youqin. Hybrid pruning algorithm for artificial neura l network training[J]. Journal of Tsinghua University, 2005,45(6):831834.
[8]JAMES T L. Statistical method of pruning neural networks[C]// International Joint Conference on Neural Networks. Washington,DC,1999.
[9]KIERON M. Fast unit selection algorithm for neural network design[C ]// 15th International Conference on Pattern Recognition. Southampton,2000.
 [10]MESSER K, KITTLER J. Choosing an optimal neural network size to a id search through a large image database[C]// Proc British Machine Vision Confe rence BMVC98.[S.l.],1998.

相似文献/References:

[1]乔俊飞,张颖.一种多层前馈神经网络的快速修剪算法[J].智能系统学报,2008,3(02):83.[doi:CNKI:SUN:ZNXT.0.2008-02-022]
 QIAO Jun-fei,ZHANG YingCollege of Electronic and Control Engineering,Beijing University of Technology,et al.Fast unit pruning algorithm for multilayer feedforward network design[J].CAAI Transactions on Intelligent Systems,2008,3(02):83.[doi:CNKI:SUN:ZNXT.0.2008-02-022]

备注/Memo

备注/Memo:
收稿日期:2007-03-22.
基金项目:
国家自然科学基金资助项目(60304012,60674066);
北京市科技新星计划资助项目(H020821210120).
作者简介:
乔俊飞,男,1968年生,教授,博士,主要研究方向为复杂过程建模与控制、计算智能与智能优化控制. 
张 颖,女,1982年生,硕士研究生,主要研究方向为污水处理过程的智能化建模与仿真 . 
通讯作者:张 颖.zhangying611@163.com.
更新日期/Last Update: 2009-05-11