Risk control model optimization method and device, computer equipment and storage medium

Abstract

The invention provides a risk control model optimization method and device, computer equipment and a storage medium. The method comprises the steps: obtaining the multi-dimensional data of a sample user, and generating the portrait data of the sample user; associating and deriving conventional characteristic factors of the portrait data of the sample user by means of a knowledge graph and a complex network according to the portrait data of the sample user; through big data, mining hidden feature factors between portrait data of the sample users; calculating a time sequence influence factor ofthe sample user; preprocessing the conventional feature factors, the hidden feature factors and the time sequence influence factors, performing feature decomposition, measuring the influence of each feature on the model accuracy, and removing noise to form machine learning model training data; and performing xgboost model training through the machine learning model training data to obtain a risk control prediction model. According to the method, a time sequence influence factor and a hidden characteristic factor are introduced, so that the model is endowed with stronger discrimination capability for good and bad samples, and erroneous judgment is reduced.

Description

technical field [0001] The present invention relates to a wind control model optimization method, device, computer equipment and storage medium, in particular to a wind control model optimization method, device, computer equipment and storage medium. Background technique [0002] At present, when Internet financial loan service companies provide loan services, they use the risk control model to review the user's loan application. However, the existing risk control model only loads some conventionally derived features, and has no user data Carrying out in-depth excavation, and at the same time, the influence of the time characteristics of sample data on model prediction is not considered during model training. There is still room for improvement in the accuracy of model prediction and the reduction of misjudgment and miskilling. Contents of the invention [0003] The technical problem to be solved by the present invention is to provide a wind control model optimization meth...

AI Technical Summary

Problems solved by technology

[0002] At present, when Internet financial loan service companies provide loan services, they use the risk control model to review the user's loan application. However, the existing risk control model only loads some conventionally derived features, and has no user data Carry out in-depth excavation, and at the same time, do not consider the impact of the time characteristics of sample data on model prediction during model training. There is still room for improvement in model prediction accuracy and reduction of misjudgment and miskilling

Images