ABSTRACT
To improve the credit evaluation ability of the Internet finance industry to loan individuals, a personal credit evaluation model is established. To reduce the influence of experimental data on the results, firstly, the independent variables are normalized by zero mean. Secondly, to solve the collinearity problem, we use principal component analysis to extract key information; Secondly, we use binomial Logit regression to test our model. The result is that compared with the Logit model, the AUC value is 0.6966, and PCA-Logit model has higher classification accuracy, and its AUC value is 0.7620. Therefore, the PCA-Logit model has a better prediction effect, and it is better than the decision tree model. In the case of high-dimensional sparse user data, it can also provide the values of each observation point in the data. It is helpful to establish a relatively more perfect personal credit evaluation system.
- ALTMAN E I. Financial ratios, discriminant analysis and the prediction of corporate banking [J]. The Journal of Finance, 1968, 23 (4): 589-609.Google ScholarCross Ref
- Xu Yanqiu, Pan Meiqin.Application of Analytic Hierarchy Process and Support Vector Machine in Personal Credit Evaluation [J]. Chinese Management Science, 2016, 24 (S1): 106-112.Google Scholar
- LIU Xiao-ya, WANG Ying-ming.Optimization of SVM personal credit evaluation model based on C4.5 algorithm [J]. Computer System Application, 2019, 28 (7): 133-138.Google Scholar
- Makowski P. Credit Scoring Branches out, Credit World [J]. 1985, 75 (7)Google Scholar
- Kass G. An Exploratory Technique for Investigating Large Questions of Categorical Data[J]. Appl Stat, 2002, 29(3)Google Scholar
- YANG Sheng-gang, ZHU Qi, CHENG Cheng. Construction of personal credit evaluation combination model–based on decision tree-neural network research [J]. Financial Forum, 2013, 18 (2): 57-61, 67.Google Scholar
- Dai Hongjun, Ji Longwei.Research on financial crisis early warning model of listed companies based on principal component and Logit [J]. Journal of Huainan Normal University, 2009 (1): 129-133Google Scholar
Recommendations
Research on Credit Card Fraud Detection Model Based on Distance Sum
JCAI '09: Proceedings of the 2009 International Joint Conference on Artificial IntelligenceAlong with increasing credit cards and growing trade volume in China, credit card fraud rises sharply. How to enhance the detection and prevention of credit card fraud becomes the focus of risk control of banks. This paper proposes a credit card fraud ...
Research on Personal Credit Evaluation System of Commercial Banks
WKDD '08: Proceedings of the First International Workshop on Knowledge Discovery and Data MiningBased on the analysis of personal credit evaluation index systems in and out of China, this paper establishes personal credit evaluation index system of commercial banks suitable for China, which can be divided into three parts: Personal indices, ...
Prognostic personal credit risk model considering censored information
Credit scoring is one of the most successful applications of quantitative analysis that helps organizations decide whether or not to grant credit to consumers who apply to them. However, standard credit risk models based on binary classifying approaches ...
Comments