User risk prediction method and system

Abstract

The invention provides a user risk prediction method and system. The method comprises: obtaining a score card according to multiple indicators; constructing prediction models online according to the score card and index conditions, constructing different prediction models according to different types of users, and constructing at least one prediction model for the same type of user; optimizing theprediction models; setting a usage ratio of each prediction model for predicting users of the same type; setting an index calculation rule for calculating an index value of each indicator; and bringing information of users into the index calculation rule to obtain the index values, bringing the obtained index values into the corresponding prediction models to obtain prediction results, and matching products that accord with the prediction results according to the prediction results and product information. The user risk prediction method and system is high in accuracy of the prediction results, and the prediction model adjustment is relatively convenient; and the user risk prediction method and system can quickly match suitable products for the users with a high pass rate, reduce the financial risk of cooperative institutions, and improve the efficiency of a user side and an institution end.

Description

technical field [0001] The invention belongs to the technical field of Internet information processing, and in particular relates to a user risk prediction method and system. Background technique [0002] Existing Internet finance, in order to reduce financial risks, will conduct risk assessments on users. Most of the existing models are not systematic, and the existing systems cannot dynamically adjust the forecasting model, and cannot flexibly build and test the model. When the index value of the index is obtained from the user information, it needs to go through the index calculation rules. The existing model is time-consuming and laborious when the index calculation rules need to be changed, and the model adjustment is troublesome. Contents of the invention [0003] The present invention is made to solve the above problems, and the purpose is to provide a user risk prediction method and system with high modeling efficiency and high accuracy, which can dynamically adju...

AI Technical Summary

Problems solved by technology

Most of the existing models are not systematic, and the existing systems cannot dynamically adjust the prediction model, and cannot flexibly build and test models

When the indicator value of the indicator is obtained from the user information, it needs to go through the indicator calculation rules. The existing model is time-consuming and laborious when the indicator calculation rules need to be changed, and the model adjustment is troublesome.

Images