Improved Weighted Support Vector Machine

被引：0

作者：

Li Wanling ^{[1
]}

Chen Peng ^{[1
]}

Song Xiangjun ^{[1
]}

机构：

[1] Troops 63908 PLA, Shijiazhuang, Peoples R China

来源：

PROCEEDINGS OF THE 2016 5TH INTERNATIONAL CONFERENCE ON ADVANCED MATERIALS AND COMPUTER SCIENCE | 2016年 / 80卷

关键词：

SVM; Weighted SVM; sample; isolated points; C-SVM;

D O I：

暂无

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

In this paper, weighted Support Vector Machine was introduced. And the weighted Support Vector Machine was improved. Some methods of determining weight were introduced, and the comprehensive method of determining weight was adopted. Simulation results indicated that the total accuracy rating by improved weighted Support Vector Machine is higher than C-SVM. It's able to improve the distribution accuracy rating with improved Supported Vector Machine effectively.

引用

页码：14 / 17

页数：4

共 5 条

[1] A comparison of methods for multiclass support vector machines [J].