[1]牛奔,李丽.基于MCPSO算法的BP神经网络训练[J].深圳大学学报理工版,2009,26(2):147-150.
 NIU Ben and LI Li.Artificial neural networks training based on MCPSO algorithm[J].Journal of Shenzhen University Science and Engineering,2009,26(2):147-150.
点击复制

基于MCPSO算法的BP神经网络训练()
分享到:

《深圳大学学报理工版》[ISSN:1000-2618/CN:44-1401/N]

卷:
第26卷
期数:
2009年2期
页码:
147-150
栏目:
电子与信息工程
出版日期:
2009-04-30

文章信息/Info

Title:
Artificial neural networks training based on MCPSO algorithm
文章编号:
1000-2618(2009)02-0147-04
作者:
牛奔李丽
深圳大学管理学院,深圳 518060
Author(s):
NIU Ben and LI Li
College of Management ,Shenzhen University ,Shenzhen 518060,P.R.China
关键词:
粒子群多群体神经网络函数逼近模式分类
Keywords:
particle swarmmulti-swarmneural networkfunction approximationpattern classification
分类号:
TP 18
文献标志码:
A
摘要:
基于多群体协同进化粒子群算法,提出一种用于BP神经网络训练的新型学习算法.将网络中需要调整权值与偏差组成的矢量看成MCPSO算法中粒子,通过粒子间的竞争与合作,完成网络训练过程.将基于 MCPSO训练的BP网络分别应用于函数逼近和模式分类问题.结果表明,基于MCPSO的神经网络学习算法在收敛速度和学习效率等方面优于其他方法.
Abstract:
A novel training algorithm based on MCPSO was proposed to train the BP neural networks was proposed.The free parameters including the weights and bias were regarded as the particles in MCPSO,and the networks were trained by competition and collaboration of the individuals in MCPSO.The designed evolutionary networks were applied to function approximation problems and pattern classification problems.The experimental results demonstrate that the MCPSO based training algorithm is superior to other training algorithms in terms of convergence rate and learning efficiency.

参考文献/References:

[1]Hornik K,Stinchcombe M,White H.多层前馈神经网络是一种通用逼近器[J].神经网络,1989,2(5):359-366 (英文版).
[2]Rumelhard D E,Hinton G E,Willianms R J.基于误差反向传播的学习表达[J].自然,1986,(323):533-536 (英文版).
[3]高海兵,高亮,周驰,等. 基于粒子群优化的神经网络训练算法研究[J].电子学报,2004,32(9) : 1572-1574.
[4]秦毅男,廖晓辉,赵庆治.一种基于粒子群优化算法的神经网络训练方法[J].河南师范大学学报自然科学版,2007,35(3) :169-171.
[5]岳琳,张宏伟,王亮.粒子群优化算法在城市需水量预测中的应用[J].天津大学学报自然科学与工程技术版,2007,40(6) :742-746.
[6]林晓梅,吕珊珊,朱丹,等.基于神经网络-粒子群优化算法的医学图像分割新方法[J].长春工业大学学报自然科学版,2008,29(2):158-161.
[7]牛奔,朱云龙,何小贤.多群体协同粒子群优化算法[C]// 2005年欧洲人工生命大会论文集.柏林:施普林格出版社,2005:874-883 (英文版).
[8]牛奔,朱云龙,何小贤,等.MCPSO:一种多群体协作粒子群优化方法[J].应用数学与计算, 2007,185(2):1050-1062 (英文版).


[1]Hornik K,Stinchcombe M,White H.Multilayer feed-forward networks are universal approximators[J].Neural Networks,1989,2(5):359-366.
[2]Rumelhard D E,Hinton G E,Willianms R J.Learning representations by back-propagation errors[J].Nature,1986,(323):533-536.
[3]GAO Hai-bing, GAO Liang,ZHOU Chi,et al.Particle swarm optimization based algorithm for neural network learning[J].Acta Electronica Sinica,2004,32(9):1572-1574 (in Chinese).
[4]QIN Yi-nan,LIAO Xiao-hui,ZHAO Qing-zhi. A neural network training algorithm based on particle swarm optimization[J].Journal of Henan Normal University (Natural Science),2007,35(3):169-171 (in Chinese).
[5]YUE Lin,ZHANG Hong-wei,WANG Liang.Application of particle swarm optimization in urban water demand prediction[J].Journal of Tianjin University:Science and Technology,2007,40(6):742-746 (in Chinese).
[6]LIN Xiao-mei,LU Shan-shan,ZHU Dan,et al.A new particle swarm optimization algorithm for medical image segmentation based on neural network[J]. Journal of Changchun University of Technology(Natural Scionce Edition),2008,29(2):158-161 (in Chinese).
[7]NIU Ben,ZHU Yun-long,HE Xiao-xian.Multi-population cooperative particle swarm optimization[C]// European Conference on Artificial Life,Springer Verlag,2005:874-883.
[8]NIU Ben,ZHU Yun-long,HE Xiao-Xian,et al.MCPSO:a multi-swarm cooperative particle swarm optimizer[J].Applied Mathematics and Computation, 2007,185(2):1050-1062.

备注/Memo

备注/Memo:
收稿日期:2008-08-25;修回日期:2009-03-16
基金项目:深港创新园资助项目(SG20081022137A)
作者简介:牛奔(1980-),男(汉族),安徽省全椒市人,深圳大学讲师、博士.E-mail:drniuben@gmail.com
更新日期/Last Update: 2009-05-15