| 回归型支撑矢量机的罚函数法及其改进算法研究 |
| |
| 引用本文： | 陈伟,胡昌华,方华元. 回归型支撑矢量机的罚函数法及其改进算法研究[J]. 电光与控制, 2005, 12(6): 52-55 |
| |
| 作者姓名： | 陈伟 胡昌华 方华元 |
| |
| 作者单位： | 第二炮兵工程学院,陕西,西安,710025;第二炮兵工程学院,陕西,西安,710025;第二炮兵工程学院,陕西,西安,710025 |
| |
| 摘 要： | 介绍了用于回归估计的支撑矢量机,针对标准罚函数法训练速度慢和不能有效区分不同样本重要程度的问题,将两种改进算法用于支撑矢量机的回归估计,并通过仿真实验分析了算法的训练效果,证明了算法的可行性。
|
| 关 键 词： | 支撑矢量机 回归估计 罚函数法 |
| 文章编号： | 1671-637X（2005）06-0052-04 |
| 收稿时间： | 2004-11-09 |
| 修稿时间： | 2004-11-092004-12-09 |
Penalty function and its improved algorithms for regression type support vector machine |
| |
| CHEN Wei,HU Chang-hua,FANG Hua-yuan. Penalty function and its improved algorithms for regression type support vector machine[J]. Electronics Optics & Control, 2005, 12(6): 52-55 |
| |
| Authors: | CHEN Wei HU Chang-hua FANG Hua-yuan |
| |
| Abstract: | The support vector machine(SVM) used for regression estimation is introduced in this paper.Since the standard penalty-function has low training speed and can't distinguish the weightiness of different samples,two improved penalty function algorithms were used for regression estimation of SVM.Simulated experiments were taken for analyzing the training effect of these algorithms.The feasibility of the algorithms was proved at last. |
| |
| Keywords: | Support Vector Machine(SVM) regression estimation penalty function algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
