Forecast of Bank Credit Risk in Fur Animal Industry

  • Han He
Keywords: Fur Animal Industry, Data Mining, Information Entropy, Decision Tree, Bank Credit

Abstract

At present, the fraudulent loans in the fur industry of the banking industry are not uncommon, and the
banking industry's risk prevention measures for this business are limited to strengthening internal management
and improving the moral quality of employees. In the face of increasingly fierce competition, these practices are
far from enough. Faced with the increasing index of the banking industry, how to effectively use data mining
technology to obtain useful information and knowledge from massive historical data, so as to effectively prevent
the credit risk of the fur animal industry, will become the future success of the banking industry in the international
competition. One of the important means. In this paper, based on the business situation, the DSB-ID3 model is
constructed by combining the classical decision tree algorithm, SMOTE sampling method, differential sampling
rate resampling technique and Bagging technique in the class unbalanced data classification processing method,
and DSB-ID3 is constructed. The model is compared with the single decision tree ID3 model and the
oversampling decision tree ID3 model to analyze its advantages and disadvantages. Experiments show that the
proposed DSB-ID3 model has significantly improved the early warning accuracy compared with the previous
models, and the model has practical application value.

Published
2019-09-01