基于最小二乘近似支持向量回归模型的电子商务信用风险预警

E-commerce credit risk early-warning with a least squares proximal support vector regression model

随着制约电子商务发展的基础性问题如交易支付和物流配送等问题的逐步解决,电子商务交易双方的信用问题已凸显为电子商务发展的瓶颈.在国内外研究现状的基础上,构建电子商务信用风险的预警指标体系,并建立电子商务信用风险预警的最小二乘近似支持向量回归模型.同时,利用相关样本数据进行了实证分析,并根据实证结果给出不同的预警管理对策.

With the rapid development of E-commerce, some fundamental issues such as payment and logistics distribution are solved gradually and E-commerce credit risk between buyers and sellers becomes a bottleneck problem restricting the development of E-commerce. For this purpose, an E-commerce credit risk early-warning indicator system is first constructed in terms of the previous studies. Then a least squares proximal support vector regression model is proposed for E-commerce credit risk early-warning. Using a public available dataset on E-commerce credit risk, empirical experiments are finally conducted. Based on some experimental results, some different early-warning management strategies are presented for E-commerce credit risk management.