| 一种新的DAGSVMs多类分类方法 |
| |
| 引用本文： | 王晓锋,温学兵,滕明岩.一种新的DAGSVMs多类分类方法[J].铁路计算机应用,2009,18(5):15-17. |
| |
| 作者姓名： | 王晓锋 温学兵 滕明岩 |
| |
| 作者单位： | 渤海大学,数学系,锦州,121013 |
| |
| 基金项目： | 辽宁省教育厅资助项目 |
| |
| 摘 要： | 分析DAGSVMs多类分类方法存在的问题,针对其存在的问题,提出一种新的DAGSVMs多类分类方法.该方法采用聚类分析中类距离的思想作为层次分类依据.实验结果表明,该方法具有较高的分类精度.
|
| 关 键 词： | 支持向量机 多类分类 DAG 方法 |
| 收稿时间： | 2009-05-15 |
New method of SVM multiclass classification based on DAG |
| |
| WANG Xiao-feng,WEN Xue-bing,TENG Ming-yan.New method of SVM multiclass classification based on DAG[J].Railway Computer Application,2009,18(5):15-17. |
| |
| Authors: | WANG Xiao-feng WEN Xue-bing TENG Ming-yan |
| |
| Institution: | ( Department of Mathematics, Bohai University, Jinzhou 121013, China ) |
| |
| Abstract: | The problems and defections of the DAG SVM multiclass classification were analyzed. To solve these problems, an new method of SVM multiclass classification based on DAG was put forward. The class distance of clustering was as the basis of hierarchical classification. The experiment results showed that the new method had higher multiclass classification accuracy than the original algorithm. |
| |
| Keywords: | DAG |
| 本文献已被 维普 万方数据 等数据库收录！ |
| 点击此处可从《铁路计算机应用》浏览原始摘要信息 |
| 点击此处可从《铁路计算机应用》下载免费的PDF全文 |
|
