Credit assessment model training method, and credit assessment method and apparatus

Abstract

The invention discloses a credit assessment model training method, and a credit assessment method and apparatus. The training method comprises the steps of obtaining original training behavior data of a training user in a business system; extracting original training features in the original training behavior data; performing feature combination on the original training features according to a GBDT model to generate corresponding training cross combination features; and performing training on a logistic regression model according to the training cross combination features to build a credit assessment model. According to the method, the original training features are trained through the nonlinear GBDT model to generate the corresponding training cross combination features, and the linear LR model is trained through the training cross combination features to build the credit assessment model, so that the credit assessment model has not only high performance of the nonlinear model but also interpretability of the linear model.

Description

technical field [0001] The present application relates to the field of data evaluation, in particular to a method for training a credit evaluation model, a method and a device for performing credit evaluation using the credit evaluation model. Background technique [0002] At present, most commonly used traditional credit information systems use scorecards to construct user credit models. Scorecards are actually a binary classification model, and the construction of this model is mainly implemented using the Logistic Regression algorithm. The scorecard model needs to preprocess the variables. Due to the need to make scores, the scorecard model will segment the variables first, and then encode the variables based on the target. The commonly used encoding method in the credit model is WOE (Weight Of Evidence) , which can be intuitively understood as the influence of a variable on the target when it takes a certain value, and this method can be used to improve the performance o...

AI Technical Summary

Problems solved by technology

[0003] However, the current problems are: in the traditional credit information system, although the use of a pure linear model can guarantee the interpretability of the model, the performance will be relatively poor, and the traditional linear model feature combination method is inefficient. As a result, the performance of the user credit model constructed is low, which may lead to inaccurate credit evaluation results for users through this model

Images