摘要
本文利用径向基网络的一种变化形式——广义回归神经网络(GRNN)提出了基于广义回归神经网络的函数逼近方法,利用matlab中的神经网络工具箱设计了GRNN模型,用于对非线性函数的逼近。通过网络的训练、测试达到了预期的效果,并与BP网络、RBF网络对比,说明GRNN网络的优势。
In this paper,author proposed function approximation methods based on generalized regression neural network,a variation of RBF network.Using the nerve network kit of matlab designed grnn model to approach of nonlinear function.Through the network training and testing achieved the desired results.Shows the advantages of GRNN network With BP network and RBF network compared.
出处
《巢湖学院学报》
2010年第6期11-16,共6页
Journal of Chaohu University
