Regional economic short-term forecasting based on LS-SVM

被引：0

作者：

Lin Jian ^{[1
]}

Zhu Bang-zhu ^{[1
]}

机构：

[1] Wuyi Univ, Inst Syst Sci & Technol, Jiangmen 529020, Peoples R China

来源：

Proceedings of 2006 Chinese Control and Decision Conference | 2006年

关键词：

regional economic; short-term forecasting; LS-SVM; BP neural network; generalization;

D O I：

暂无

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

A new methodology based on least squares support vector machine (LS-SVM) for regional economic short-term forecasting is presented. Due to the constraints of inequalities of the classic SVM replaced by equality-type constraints in LS-SVM, the solution follows directly from solving a set of linear equations instead of quadratic programming. Then regional economic short-term forecasting model based on LS-SVM is proposed and applied to predict GDP of Jiangmen The results show, that LS-SVM has excellent learning Ability and generalization, with can achieve greater accuracy than BP neural network with only few samples.

引用

页码：913 / 916

页数：4