Credit Risk Assessment Of Internet Education Based On Statistical Learning

Since 2010,there has been an upsurge of Internet Education and entrepreneurship in China.Compared with traditional offline education,internet education breaks time and space constraints,comprehensively integrates excellent teaching resources,and further promotes "educational equity".In December 2019,COVID-19 broke out,and online learning became a new normal.Internet education industry ushered in a new wave of investment boom.Under the attractive development prospect,internet education enterprises are facing the entrepreneurial risk of difficult to attract users and low survival rate.In order to survive and make profits in the changeable and complex market environment,enterprises need to improve their risk prevention ability,and the credit risk assessment of Internet education enterprises is particularly important.Considering the differences of business conditions and business models in different industries,this paper establishes a credit evaluation model in line with the Internet education industry.First of all,it selects 24 indicators of the enterprise’s credit structure and operating capacity to conduct a preliminary analysis from the six dimensions of the enterprise’s credit structure and operating capacity.Secondly,taking 117 internet education enterprises as samples and 4 enterprises specially treated as samples with serious credit risk,the data are preprocessed and smote algorithm is used to deal with the problem of sample imbalance.Then the significance test is carried out.Thirdly,21 indicators are retained,the credit risk evaluation index system of Internet education enterprises is established,and the characteristic variables are extracted by factor analysis.Finally,establish the model: 1)two classification model,establish logistic model and random forest model,and evaluate the model with AUC value and stability coefficient;2)The grading model establishes the credit rating model of Internet education enterprises based on K-means algorithm,stipulates that the credit rating of enterprises is divided into six levels,and calculates the limit of credit rating.Based on the grading model,credit rating is carried out for specific enterprises,and the changes of credit rating of Internet education enterprises are dynamically analyzed to provide suggestions for risk prevention of Internet education enterprises.