| 一种支持向量回归的局部邻域稀疏化方法 |
| |
| 引用本文： | 赵汗青.一种支持向量回归的局部邻域稀疏化方法[J].火力与指挥控制,2008(Z2). |
| |
| 作者姓名： | 赵汗青 |
| |
| 作者单位： | 装甲兵工程学院; |
| |
| 摘 要： | 调整高斯核函数参数可以改变其VC维,通过密度聚类算法发现并分离全局中非线性复杂度不同的局部特征,以支持向量比率为优化指标确定满足局部稀疏条件的核函数,从而达到优化核函数选择以提高整体回归稀疏性的目的。
|
| 关 键 词： | 支持向量回归 稀疏性 核密度聚类 |
A Localized Sparseness Regression for Support Vector Machine |
| |
| ZHAO Han-qing.A Localized Sparseness Regression for Support Vector Machine[J].Fire Control & Command Control,2008(Z2). |
| |
| Authors: | ZHAO Han-qing |
| |
| Institution: | ZHAO Han-qing(Academy of Armored Forces Engineering,Beijing 100072,China) |
| |
| Abstract: | The VC dimension of Gaussian Kernel is determined by its width parameter.The local nonlinear properties can be found and isolated by density based clustering algorithm.A proper Gaussian Kernel parameter can be obtained by optimizing a support vector number ratio.The sparseness in support vector regression is finally approved by this algorithm. |
| |
| Keywords: | Support Vector Regression(SVR) sparseness kernel density clustering |
| 本文献已被 CNKI 等数据库收录！ |
