Lin Xue,  Yuhjen Cho, Liujing Yao, and Wei He.  Model and Application of College Students Credit Evaluation Based on Logistic Regression.  Dynamic Systems and Applications 29 (2020) No. 3, 907-922

https://doi.org/10.46719/dsa202029341

ABSTRACT.
Starting from the value correspondence of big data technology application and college students credit risk management, based on the college students credit risk evaluation index system including nine input variables, one target variable and one weight variable, this paper selects 3000 students at random as the modeling samples to establish a credit evaluation model combined with five steps including determining modeling data variables, inputting variable binning, establishing a Logistic regression model of the input variables and the target variable, converting Logistic regression coefficient into credit score, and testing the fitting effect of the model. The empirical results show that the scores of the binning of the variables are easy to understand and monitor, and when the credit scores equal to 516, the Kolmogorov-Smirnov index value reaches 29.067%, the model fitting effect is acceptable. High-efficiency functioning of college students credit risk management pattern, the value standpoint on the potential obstacles of privacy difficulty and multi-source data collection difficulty can reduce the risk of technology and system validity.

Keywords: college students; credit risk evaluation index system; credit evaluation model; Logistic regression; credit score; credit risk management pattern; potential obstacles