[1]刘春林,何建敏.神经网络用于模式识别分类的改进算法[J].东南大学学报(自然科学版),1999,29(1):20-24.[doi:10.3969/j.issn.1001-0505.1999.01.004]
 Liu Chunlin,He jianmin.An Improved ANN Algorithm for Pattern Recognition[J].Journal of Southeast University (Natural Science Edition),1999,29(1):20-24.[doi:10.3969/j.issn.1001-0505.1999.01.004]
点击复制

神经网络用于模式识别分类的改进算法()
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1期
页码:
20-24
栏目:
数学、物理学、力学
出版日期:
1999-01-20

文章信息/Info

Title:
An Improved ANN Algorithm for Pattern Recognition
作者:
刘春林 何建敏
东南大学经济管理学院,南京 210096
Author(s):
Liu Chunlin He jianmin
School of Economics and Management, Southeast University, Nanjing 210096
关键词:
BP算法 模式识别 距离
Keywords:
BP algorithm pattern recognition distance
分类号:
O221.2
DOI:
10.3969/j.issn.1001-0505.1999.01.004
摘要:
用BP神经网络算法进行模式识别分类,即使对一个训练比较好的网络也极有可能出现样本的导师模式(真实模式)与网络判定模式不符的情况,即会出现误判样本. 当待判样本与某一误判训练样本比较接近时,网络很可能对其造成模式误判. 为此,本文通过引入训练样本的正、误判子集及定义在其上的待判样本的距离,将距离算法和BP算法相结合,提出了解决这一问题的新方法.
Abstract:
BP algorithm applied in pattern recognition can cause some learned specimens (samples) to be classified erroneously, even if the neural network is trained well. When a new input to be evaluated is very close to one of these misclassified specimens, the pattern of the input can also be erroneously identified by ANN. By introducing the concept of distance and two subsets (error training set and correct training set) included in the training set, this paper gives a new method which makes BP algorithm and distance algorithm integrated.

参考文献/References:

[1] Yu Xiaohu.Dynamic learning rate optimization of the backpropagation algorithm.IEEE Trans on Neural Networks,1996,6(3):669~677
[2] Palos A.An accelerated learning algorithm for multilayer perceptron networks,IEEE Trans on Neural Networks,1994,5(3):493~497
[3] Timo Sorga.Neural networks.In process fault diagnosis.IEEE Trans on SMC,1991,21(4):815~825
[4] Becraft W R,Lee P L.An integrated neural networks/expert system approach for fault diagnosis.Computers in Industry,1991,16:169~177
[5] 方开泰.实用多元统计分析.上海:华东师范大学出版社,1989.179~256
[6] 何亦华.人工神经网络用于有机化合物定量结构——活性关系研究.见:全国96神经网络学术会议,上海:1996

相似文献/References:

[1]袁浩,薛澄歧,卢章平.工业产品数字化模型的检索方法[J].东南大学学报(自然科学版),2006,36(3):397.[doi:10.3969/j.issn.1001-0505.2006.03.013]
 Yuan Hao,Xue Chengqi,Lu Zhangping.Digital model retrieval for industrial products[J].Journal of Southeast University (Natural Science Edition),2006,36(1):397.[doi:10.3969/j.issn.1001-0505.2006.03.013]
[2]王强,达飞鹏,宋文忠.前向神经网络临时极小点动态性能[J].东南大学学报(自然科学版),2005,35(4):641.[doi:10.3969/j.issn.1001-0505.2005.04.032]
 Wang Qiang,Da Feipeng,Song Wenzhong.Dynamics of temporary minima in feed-forward neural network[J].Journal of Southeast University (Natural Science Edition),2005,35(1):641.[doi:10.3969/j.issn.1001-0505.2005.04.032]
[3]程力,蔡体菁.基于模式识别神经网络的重力匹配算法[J].东南大学学报(自然科学版),2007,37(5):839.[doi:10.3969/j.issn.1001-0505.2007.05.020]
 Cheng Li,Cai Tijing.Gravity matching algorithm based on pattern recognition neural network[J].Journal of Southeast University (Natural Science Edition),2007,37(1):839.[doi:10.3969/j.issn.1001-0505.2007.05.020]
[4]金世俊,黄惟一.基于多元统计分析的触觉图象模式识别[J].东南大学学报(自然科学版),1996,26(5):95.[doi:10.3969/j.issn.1001-0505.1996.05.018]
 Jin Shijun,Huang Weiyi.Tactile Image Pattern Recognition Based on Multivariate Statistical Analysis[J].Journal of Southeast University (Natural Science Edition),1996,26(1):95.[doi:10.3969/j.issn.1001-0505.1996.05.018]
[5]李涛,胡爱群,高尚.基于BB-BM算法的网络协议内容符合性测试方法[J].东南大学学报(自然科学版),2015,45(6):1027.[doi:10.3969/j.issn.1001-0505.2015.06.001]
 Li Tao,Hu Aiqun,Gao Shang.Context consistency test method for network protocol based on BB-BM algorithm[J].Journal of Southeast University (Natural Science Edition),2015,45(1):1027.[doi:10.3969/j.issn.1001-0505.2015.06.001]

备注/Memo

备注/Memo:
第一作者:男,1970年生,博士研究生.
更新日期/Last Update: 1999-01-20