基于粒子群优化的过程神经网络学习算法
|
摘要
基于粒子群优化为过程神经元网络提出了一种新的学习算法。新算法在对网络输入函数和连接权函数进行正交基函数展开后,将网络中的结构参数和其他参数整合成一个粒子,再用粒子群优化算法进行全局优化。新算法不依赖于函数梯度信息,不需要手动调节网络结构。粒子群优化具有良好的全局优化性能和收敛性能,保证了过程神经元网络的全局学习能力和新学习算法的收敛能力,更好地发挥过程神经网络的逼近性能。两个实际预测问题的实验结果表明,基于粒子群优化的学习算法比现有的基于梯度的基函数展开方法以及误差反传神经网络模型具有更好的预测精度。
This paper proposes a new learning algorithm for process neural networks(PNNs) based on particle swarm optimization(PSO),called PSO-LM.After the orthogonal basis function expansion to the input functions and the weight functions of the PNN,the structure parameters and other parameters in the PNN will be formed as a particle,and globally optimized by PSO.This algorithm does not need any gradient calculations or the manual control of the network's structure.The global learning capability and the convergence capability of the PNN can be guaranteed by the capabilities of PSO,so the PSO-LM can better develop and improve the approximation capability of the PNN.According to two practical prediction applications,PSO-LM can outperform the existing basis function expansion based learning algorithm(BFE-LM) for PNNs,and the classic back propagation neural networks(BPNNs) on predictive accuracy.
This paper proposes a new learning algorithm for process neural networks(PNNs) based on particle swarm optimization(PSO),called PSO-LM.After the orthogonal basis function expansion to the input functions and the weight functions of the PNN,the structure parameters and other parameters in the PNN will be formed as a particle,and globally optimized by PSO.This algorithm does not need any gradient calculations or the manual control of the network's structure.The global learning capability and the convergence capability of the PNN can be guaranteed by the capabilities of PSO,so the PSO-LM can better develop and improve the approximation capability of the PNN.According to two practical prediction applications,PSO-LM can outperform the existing basis function expansion based learning algorithm(BFE-LM) for PNNs,and the classic back propagation neural networks(BPNNs) on predictive accuracy.
引文
[1]何新贵,许少华.过程神经元网络.北京:科学出版社,2007
[2]Kennedy J,Eberhart R C.Particle swarm optimization//Proceedings of the IEEE International Conference onNeural Networks.Perth,Western Australia,1995:1942-1948
[3]Ismail A,Engelbrecht A P.Training product unit infeedforward neural networks using particle swarmoptimization//Proceedings of the InternationalConference on Artificial Intelligence.Stockholm,1999:36-40
[4]Bergh V D F,Engelbrecht A P.Cooperative learning inneural networks using particle swarm optimizers.SouthAfrican Computer Journal,2000,26:84-90
[5]Eberhart R C,Shi Yuhui.Particle swarm optimization:developments,applications and resources//Proceedingsof the IEEE Congress on Evolutionary Computation.Seoul,2001,1:81-86
[6]Tan Ying,Xiao Zhongmin.Clonal particle swarmoptimization and its applications//IEEE Congress onEvolutionary Computation,Singapore,2007:2303-2309
[7]Zhang Junqi,Liu Kun,Tan Ying,et al.Magnifierparticle swarm optimization for numerical optimization//Proceedings of Genetic and Evolutionary ComputationConference.Atlanta,2008:167-168
[8]Zhang Junqi,Liu Kun,Tan Ying,et al.Random blackhole particle swarm optimization and its application//International Conference on Neural Networks and SignalProcessing.Zhenjiang,2008:359-365
[9]许少华,何新贵.基于函数正交基展开的过程神经网络学习算法.计算机学报,2004,27(5):645-650
[10]Wu Tianshu,Xie Kunqing,Song Guojie,et al.Numerical learning method for process neural network//International Symposium on Neural Networks.Wuhan,2009:670-678
[11]Bratton D,Kennedy J.Defining a standard for particleswarm optimization//Proceedings of the IEEE SwarmIntelligence Symposium.Hawaii,2007:120-127
[12]国家地震科学数据共享中心.地形变数据库[DB/OL].(2002-01-01)[2010-03-20].http://data-prc.enc.earthquake.cn/qzsj/d_mtd_060.htm
[2]Kennedy J,Eberhart R C.Particle swarm optimization//Proceedings of the IEEE International Conference onNeural Networks.Perth,Western Australia,1995:1942-1948
[3]Ismail A,Engelbrecht A P.Training product unit infeedforward neural networks using particle swarmoptimization//Proceedings of the InternationalConference on Artificial Intelligence.Stockholm,1999:36-40
[4]Bergh V D F,Engelbrecht A P.Cooperative learning inneural networks using particle swarm optimizers.SouthAfrican Computer Journal,2000,26:84-90
[5]Eberhart R C,Shi Yuhui.Particle swarm optimization:developments,applications and resources//Proceedingsof the IEEE Congress on Evolutionary Computation.Seoul,2001,1:81-86
[6]Tan Ying,Xiao Zhongmin.Clonal particle swarmoptimization and its applications//IEEE Congress onEvolutionary Computation,Singapore,2007:2303-2309
[7]Zhang Junqi,Liu Kun,Tan Ying,et al.Magnifierparticle swarm optimization for numerical optimization//Proceedings of Genetic and Evolutionary ComputationConference.Atlanta,2008:167-168
[8]Zhang Junqi,Liu Kun,Tan Ying,et al.Random blackhole particle swarm optimization and its application//International Conference on Neural Networks and SignalProcessing.Zhenjiang,2008:359-365
[9]许少华,何新贵.基于函数正交基展开的过程神经网络学习算法.计算机学报,2004,27(5):645-650
[10]Wu Tianshu,Xie Kunqing,Song Guojie,et al.Numerical learning method for process neural network//International Symposium on Neural Networks.Wuhan,2009:670-678
[11]Bratton D,Kennedy J.Defining a standard for particleswarm optimization//Proceedings of the IEEE SwarmIntelligence Symposium.Hawaii,2007:120-127
[12]国家地震科学数据共享中心.地形变数据库[DB/OL].(2002-01-01)[2010-03-20].http://data-prc.enc.earthquake.cn/qzsj/d_mtd_060.htm
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心