| 用FSVM解决多类分类问题中不可分区域问题剖析 |
| |
| 引用本文： | 徐文彬,陈美珠.用FSVM解决多类分类问题中不可分区域问题剖析[J].常熟理工学院学报,2007,21(8):84-88. |
| |
| 作者姓名： | 徐文彬 陈美珠 |
| |
| 作者单位： | 1. 常熟理工学院,计算机科学与工程系,江苏,常熟,215500
2. 泰州职业技术学院,江苏,泰州,225300 |
| |
| 摘 要： | 分析了支持向量机(SVM)的工作原理和将其推广到多类分类时会遇到的问题,对用模糊SVM(FSVM)解决此问题时的模糊策略作了详细论证,说明此模糊策略是非常完美的一个解决方案,并指出了进一步的研究方向。
|
| 关 键 词： | 支持向量机 模糊支持向量机 分类超平面 |
| 文章编号： | 1008-2794（2007）08-0084-05 |
| 收稿时间： | 2007-04-10 |
An Analysis of Resolving the Problem of Unclassifiable Regions in Multi-classification by Means of FSVM |
| |
| XU Wen-bin,CHEN Mei-zhu.An Analysis of Resolving the Problem of Unclassifiable Regions in Multi-classification by Means of FSVM[J].Journal of Changshu Institute of Technology,2007,21(8):84-88. |
| |
| Authors: | XU Wen-bin CHEN Mei-zhu |
| |
| Abstract: | This paper analyzes the principle of support vector machine(SVM) and points out problems encountered in its extension to multi-classification.To solve this problem,fuzzy strategy of fuzzy SVM(FSVM) has been demonstrated in detail.And this vague strategy is a very good solution.The paper also suggests a direction for further research. |
| |
| Keywords: | SVM FSVM classification hyperplane |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
