基于量子差分进化算法的神经网络优化方法

被引:10
作者
杜文莉
周仁
赵亮
钱锋
机构
[1] 华东理工大学化工过程先进控制和优化技术教育部重点实验室
关键词
神经网络; 差分进化算法; 协同量子差分进化算法(CQGADE);
D O I
10.16511/j.cnki.qhdxxb.2012.03.014
中图分类号
TP183 [人工神经网络与计算];
学科分类号
摘要
一般的神经网络的结构是固定的,在实际应用中容易造成冗余连接和高计算成本。该文采用了协同量子差分进化算法(cooperative quantum differential evolution algo-rithm,CQGADE)以同时优化神经网络的结构和参数,即采用量子遗传算法(quantum genetic algorithm,QGA)来优化神经网络的结构和隐层节点数,采用差分算法来优化神经网络的权值。训练后的神经网络的连接开关能有效删除冗余连接,算法的量子概率幅编码和协同机制可以提高神经网络的学习效率、逼近精度和泛化能力。仿真实验结果表明:用训练后的神经网络预测太阳黑子和蒸汽透平流量具有更好的预测精度和鲁棒性。
引用
收藏
页码:331 / 335
页数:5
相关论文
共 12 条
[1]  
Training of artificial neural net-works using differential evolutional algorithm. Adam Slowik,Michal Bialko. 2008 Conferenceon Human System Interaction . 2008
[2]  
Tuning the structure and parameters of a neural network using an orthogonal simulated annealing algorithm. Li-Sun Shu,shinn-Ying Ho,Shinn-Jang Ho. 2009 Joint conferences on Pervasive Computing . 2009
[3]  
Tuning of the structure and parameters of neural network using an improved genetic algorithm. Leung H. F.,Lam H. K.,Ling S. H.,Tam K. S. Proc.27th Annu. Conf. IEEE Industrial Electronics Society . 2001
[4]  
Application of BP neural network for line losses calculationbased on quantum genetic algorithm. LIU Kewen,ZHOU Haiming,YANG Zhanzhong,et al. 2011 4thInternational Symposium on Computational Intelligence andDesign . 2011
[5]  
The summary of differentialevolution algorithm and its improvement. LI Gaoyang,LIU Minguang. 3rdInternational Conference on Advanced Computer Theory andEngineering . 2010
[6]  
Tuning the structure and parameters of aneural network by using hybrid Taguchi-genetic algorithm. Tsai J,Chou J,Liu T. IEEE Transaction on Neural Network . 2006
[7]  
Tuning ofthe structure and parameters of a neural network using agoodpoints set evolutionary strategy. XIAO Chixin,CAI Zixing,WANG Yong,et al. The 9th Internationalconference for Young Computer Scientists . 2008
[8]  
Differential Evolution – A Simple and Efficient Heuristic for global Optimization over Continuous Spaces[J] . Rainer Storn,Kenneth Price. &nbspJournal of Global Optimization . 1997 (4)
[9]  
基于协同PSO算法的模糊辨识与神经网络学习[D]. 赵亮.上海交通大学 2009
[10]   量子进化算法 [J].
杨淑媛 ;
焦李成 ;
刘芳 .
工程数学学报, 2006, (02) :235-246