基于KPCA-GaussianNB的电子商务信用风险分类

E-commerce Credit Risk Classification Based on KPCA-Gaussian NB

用核主成分分析(KPCA)和高斯朴素贝叶斯(GaussianNB)构建电子商务信用风险分类模型(KPCAGaussianNB)。首先,通过KPCA方法将电子商务信用风险涉及的指标进行主要特征提取;其次,应用GaussianNB方法构造电子商务信用风险分类模型;最后,使用18家电子商务企业的真实数据进行实证检验,并依据检验结果提出应对风险的措施。验证结果表明:通过对比GaussianNB、PCA-GaussianNB和KPCA-GaussianNB的平均准确率,KPCA-GaussianNB的平均准确率最高。

In this paper,an e-commerce credit risk classification model(KPCA-Gaussian NB) is constructed using the kernel principal component analysis(KPCA) and Gaussian NB.Firstly,the KPCA method is used to extract the main features of the indexes involved with the e-commerce credit risk.Secondly,the Gaussian NB method is used to construct the e-commerce credit risk classification model.Finally,the empirical data of 18 e-commerce enterprises are used to test the effectiveness of the model,and based on the test results,to propose the corresponding measures to deal with these risks.Through comparison,the KPCA-Gaussian NB is found to have higher average accuracy than Gaussian NB and PCA-Gaussian NB.