计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (22): 224-226.

• 工程与应用 • 上一篇    下一篇

基于OLS与EPSO算法的RBF企业订单预测模型研究

宫蓉蓉   

  1. 长沙民政职业技术学院,长沙 410004
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-08-01 发布日期:2011-08-01

Research of business order forecast model based on radial basis function neural network using OLS and EPSO algorithms

GONG Rongrong   

  1. Changsha Social Work College,Changsha 410004,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-08-01 Published:2011-08-01

摘要: 提出了一种最小正交二乘算法(OLS)和进化粒子群优化算法(EPSO)相结合构建RBF神经网络的企业订单预测模型。OLS采用前向回归算法,从输入数据中选取适当的中心,动态地避免网络规模过大和随机选择中心带来的数值病态问题;EPSO方法调整网络中的参数,如RBF中心位置,RBF宽度和隐层与输出层之间的权值,以提高网络的泛化能力。

关键词: 径向基函数(RBF), 最小正交二乘算法(OLS), 进化粒子群优化算法(EPSO), 订单预测

Abstract: This paper presents a method for designing business order forecast model based on combination of Orthogonal Least Squares algorithm(OLS) and Evolutionary Particle Swarm Optimization algorithm(EPSO).OLS selects a suitable set of centers from the input data and adopts forward regression procedure to avoid bigger size of networks and numerical morbidity problem by selecting center randomly.The EPSO approach is used to tune the parameters in the network,including the position of RBF centers,the width of RBFs,and the weighting values between hidden and output layer,which improve the generalization ability of the network.

Key words: Radial Basis Function Neural Network(RBF), Orthogonal Least Squares algorithm(OLS), Evolutionary Particle Swarm Optimization algorithm(EPSO), order forecast