-
About Us+
-
Products+
-
Solution+
-
Cases+
-
AI Lab+
-
Investor+
-
Contact+
-
Join Us +
-
-
Trial+
About Us+
Products+
Solution+
Cases+
AI Lab+
Investor+
Contact+
Join Us +
Trial+
This paper proposes a set of process methods to analyze user credit data to establish credit rating according to some characteristics of banking industry. Firstly, using smote over sampling method to adjust the unbalanced data, adjusting the original unbalanced data to balanced, and using AUC value and substitution price sensitive error rate as the index of evaluating the model; Finally, the classification data of users are not only predicted: in order to ensure the accuracy of the model and meet the characteristics of large amount of data in financial industry, the training data of xbboost algorithm is used, and the user credit situation is further analyzed by combining the probability of the user category reported by the model and the knowledge of the scoring card.
Download