计算机技术与发展2015，Vol.25Issue(11)：38-43,48,7.DOI:10.3969/j.issn.1673-629X.2015.11.008
面向不平衡数据的模糊支持向量机
Fuzzy Support Vector Machine for Imbalanced Data
摘要
Abstract
Traditional Fuzzy Support Vector Machines (FSVM) are sensitive to imbalanced data. They compute their fuzzy memberships mainly according to the factor of distance,which can not reflect the importance of the samples precisely and may lead to an error of classi-fication results. To these problems,an improved FSVM is proposed in this paper. In the proposed FSVM,samples are firstly separated into different categories based on sample densities, and then they are assigned different fuzzy memberships. This method may improve the weight of support vectors and reduce the influence of outlier and noise points. Furthermore,the imbalanced factor is introduced to improve the classification precision of imbalanced data. The experimental results show that the improved FSVM has better performance for imbal-anced data with more outlier and noise points.关键词
支持向量机/模糊支持向量机/不平衡数据集/样本密度Key words
support vector machine/FSVM/imbalanced data/sample density分类
信息技术与安全科学引用本文复制引用
刘凌,郭剑,韩崇..面向不平衡数据的模糊支持向量机[J].计算机技术与发展,2015,25(11):38-43,48,7.基金项目
国家自然科学基金资助项目(61171053,61300239) （61171053,61300239）
教育部博士点基金资助项目(20113223110002) （20113223110002）
中国博士后科学基金资助项目(2014M551635) （2014M551635）
江苏省博士后科研资助计划项目(1302085B) （1302085B）
