| 基于DE-SVM非线性组合预测模型的研究 |
| |
| 引用本文： | 陈涛.基于DE-SVM非线性组合预测模型的研究[J].计算机工程与应用,2011,47(13):33-36. |
| |
| 作者姓名： | 陈涛 |
| |
| 作者单位： | 陕西理工学院 数学系，陕西 汉中 723000 |
| |
| 基金项目： | 陕西省教育厅自然科学研究项目 |
| |
| 摘 要： | 为提高预测精度，提出DE-SVM非线性组合预测模型。以组合预测模型的误差平方和最小为优化准则，用差分进化算法（DE）对支持向量机参数进行优化，并利用支持向量机对单一模型的预测结果进行组合预测。算例结果表明，DE-SVM综合利用了各单个预测模型的重要预测信息，其预测误差远远小于各单个模型的预测误差，其预测精度更高，模型的实用性更强。
|
| 关 键 词： | 支持向量机 差分进化算法 非线性 组合预测 |
| 修稿时间： | |
Research on nonlinear combined forecasting model based on DE-SVM |
| |
| CHEN Tao.Research on nonlinear combined forecasting model based on DE-SVM[J].Computer Engineering and Applications,2011,47(13):33-36. |
| |
| Authors: | CHEN Tao |
| |
| Affiliation: | Department of Mathematics，Shaanxi University of Technology，Hanzhong，Shaanxi 723000，China |
| |
| Abstract: | A nonlinear combined forecasting model based on DE-SVM is presented to improve the accuracy of forecasting,where optimal rule is the least sum of square error of the combined forecasting model,and the parameters of SVM are optimized by using DE,and the results of each forecasting model are forecasted by using SVM.The simulation results show that DE-SVM can synthesize important forecasting information of each forecasting model.Its forecasting error is obviously less than that of each model.The accuracy of forecasting will be higher and practicality of the model will be stronger |
| |
| Keywords: | Support Vector Machine （SVM） Differential Evolution （DE） nonlinear combined forecast |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《计算机工程与应用》浏览原始摘要信息 |
|
点击此处可从《计算机工程与应用》下载全文 |
