基于粗糙集和支持向量机的电子商务信用风险分类  

E-business Credit Risk Classification Based on Rough Set,Genetic Algorithm and Support Vector Machine

作　　者：邬建平 

机构地区：[1]岭南师范学院商学院,广东湛江524048

出　　处：《数学的实践与认识》2016年第13期87-92,共6页Mathematics in Practice and Theory

基　　金：国家级星火计划项目:农村信息化村级站点孵化平台建设(2013GA780086)

摘　　要：准确识别电子商务信用风险,有利于提高企业风险防范能力,减少损失.建立了基于粗糙集(RS)、遗传算法(GA)和支持向量机(SVM)的电子商务信用风险分类模型(RS-GA-SVM).首先,利用RS对分类指标进行约简,选择出电子商务信用风险关键影响因素.其次,采用GA算法优化SVM模型参数,并应于电子商务信用风险分类.最后,实证表明,RS-GA-SVM模型具有高的分类精度和分类效率.Accurate identification of e-commerce credit risk classification is helpful for improving capacity of enterprise risk prevention, so as to reduce the loss. This paper proposes an integrated model （RS-GA-SVM） of rough set （RS）, genetic algorithm （GA） and support vector machine （SVM） for E-commerce credit risk classification. Firstly, RS is used to choose the key factors of E-commerce credit risk classification. Secondly, the parameters of SVM model are optimized by using GA algorithm, which is used for E-commerce credit risk classification. Finally, the empirical results show that the proposed RS-GA-SVM model has higher classification accuracy and classification efficiency.

关 键 词：电子商务信用风险分类 粗造集 支持向量机 遗传算法 

分 类 号：F724.6[经济管理—产业经济]