| 基于十折法的最小二乘支持向量机参数选取方法 |
| |
| 引用本文： | 杨敬娜.基于十折法的最小二乘支持向量机参数选取方法[J].机械工程师,2011(12):28-29. |
| |
| 作者姓名： | 杨敬娜 |
| |
| 作者单位： | 河北软件职业技术学院智能工程系,河北保定,071000 |
| |
| 摘 要： | 针对最小二乘支持向量机参数选取的问题,提出一种基于十折法的参数选取方法.将数据平均分成十组,采用一组作为训练特征,其它作为测试组,规定参数的可能范围,然后选取平方差最小的作为参数的取值.经实验证明结果令人满意,提出的方法是可行的.
|
| 关 键 词： | 最小二乘支持向量机 参数选取 十折法 |
Parameters Selection Method Based on Ten-fold Least Squares Support Vector Machines |
| |
| YANG Jing-na.Parameters Selection Method Based on Ten-fold Least Squares Support Vector Machines[J].Mechanical Engineer,2011(12):28-29. |
| |
| Authors: | YANG Jing-na |
| |
| Affiliation: | YANG Jing-na (Department of Intelligent Engineering, Hebei Software Institute, Baoding 071000, China) |
| |
| Abstract: | Parameters selection method based on ten-fold least squares support vector machines is presented. The data are divided into ten groups, using a set of characteristics as the training, the other as the test group, the provisions of the parameters possible, and then the smallest square difference value is selected as a parameter. The experimental results is satisfactory. The proposed method is feasible. |
| |
| Keywords: | least squares support vector machines parameters selecting ten-fold method |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
