| Weighted Proximal Support Vector Machines： Robust Classification |
| |
| 引用本文： | ZHANGMeng FULi-hua WANGGao-feng HUJi-cheng. Weighted Proximal Support Vector Machines： Robust Classification[J]. 武汉大学学报:自然科学英文版, 2005, 10(3): 507-510. DOI: 10.1007/BF02831134 |
| |
| 作者姓名： | ZHANGMeng FULi-hua WANGGao-feng HUJi-cheng |
| |
| 作者单位： | [1]SchoolofComputer,WuhanUniversity,Wuhan430072,Hubei,China／／AcademyofMicroelectronicsandInformationTechnology,WuhanUniversity,Wuhan430072,Hubei,China [3]DepartmentofMathematicsandPhysics,ChinaUniversityofGeosciences,Wuhan430074,Hubei,China |
| |
| 基金项目： | SupportedbytheNationalNaturalScienceFoun dationofChina(90307017,60376031,60444004)andtheResearch FoundationforOutstandingYoungTeachers,ChinaUniversityofGe osciences(Wuhan)(CUGQNL0520) |
| |
| 摘 要： | Despite of its great efficiency for pattern classification, proximal support vector machines (PSVM), a new version of SVM proposed recently, is sensitive to noise and outliers. To overcome the drawback, this paper modifies PSVM by associating a weight value with each input data of PSVM. The distance between each data point and the center of corresponding class is used to calculate the weight value. In this way, the effect of noise is reduced. The experiments indicate that new SVM, weighted proximal support vector machine (WPSVM), is much more robust to noise than PSVM without loss of computationally attractive feature of PSVM.
|
| 关 键 词： | 数据分类 线性方程 PSVM 支撑向量装置 |
| 收稿时间： | 2004-06-03 |
Weighted proximal support vector machines: Robust classification |
| |
| Zhang Meng,Fu Li-hua,Wang Gao-feng,Hu Ji-cheng. Weighted proximal support vector machines: Robust classification[J]. Wuhan University Journal of Natural Sciences, 2005, 10(3): 507-510. DOI: 10.1007/BF02831134 |
| |
| Authors: | Zhang Meng Fu Li-hua Wang Gao-feng Hu Ji-cheng |
| |
| Affiliation: | (1) School of Computer, Wuhan University, 430072 Wuhan Hubei, China;(2) Academy of Microelectronics and Information Technology, Wuhan University, 430072 Wuhan Hubei, China;(3) Department of Mathematics and Physics, China University of Geosciences, 430074 Wuhan Hubei, China |
| |
| Abstract: | Despite of its great efficiency for pattern classification, proximal support vector machines (PSVM), a new version of SVM proposed recently, is sensitive to noise and outliers. To overcome the drawback, this paper modifies PSVM by associating a weight value with each input data of PSVM. The distance between each data point and the center of corresponding class is used to calculate the weight value. In this way, the effect of noise is reduced. The experiments indicate that new SVM, weighted proximal support vector machine (WPSVM), is much more robust to noise than PSVM without loss of computationally attractive feature of PSVM. |
| |
| Keywords: | data classification support vector machines linear equation |
| 本文献已被 CNKI 维普 万方数据 SpringerLink 等数据库收录！ |
|
