Computer Engineering and Applications ›› 2007, Vol. 43 ›› Issue (19): 177-179.

• 数据库与信息处理 • Previous Articles     Next Articles

Forecasting method of temporal data based on support vector regress machine

ZHUANG Bin,MENG Zhi-qing   

  1. College of Business and Administration,Zhejiang University of Technology,Hangzhou 310032,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-07-01 Published:2007-07-01
  • Contact: ZHUANG Bin

基于支持向量机的时态数据预测方法

庄 彬,孟志青   

  1. 浙江工业大学 经贸管理学院,杭州 310032
  • 通讯作者: 庄 彬

Abstract: Support Vector Regress machine(SVR) will be a promising method in temporal data forecasting fields because it uses a risk function consisting of the empirical error and a regularized term which is derived from the structural risk minimization principle.This paper briefly introduces the basic theory of Support Vector Regress(SVR) and applies SVR to create a model,which also can be used for forecasting the multi-attribute temporal data and the temporal data.The result of simulation shows that SVR is superior to BP Neutral Network in the stability and accuracy.

摘要: 支持向量回归机使用由经验误差项和常数项所构成的风险函数,满足结构风险最小原则。在时态数据预测领域,它将成为一种很有前途的预测方法。简要介绍了回归支持向量机的基本理论。基于回归支持向量机模型,建立了一个对时态数据预测的方法,可以对多属性时态数据进行预测,并与其它预测模型(BP神经网络)进行比较。实验结果表明所提出的方法在预测的稳定性和准确性方面都要优于BP神经网络模型。