Risk control model training method and device, and equipment

Abstract

The invention discloses a risk control model training method and device, and equipment. When the black samples in the training samples are impure and the risk control model is trained, cross validation is carried out on the training samples through other related trained risk identification models, so that the black samples with low scores and the white samples with high scores are filtered out, and the model is retrained by utilizing the screened samples.

Description

technical field [0001] The embodiments of this specification relate to the field of information technology, and in particular to a method, device and equipment for training a risk control model. Background technique [0002] In the process of model training, there are many training sample sets that contain enough white samples, but lack enough black samples, or the black samples are not pure enough. [0003] For example, in the field of risk control, users or merchants use their accounts to conduct illegal / non-compliant behaviors to generate active risks. The common ones are spam registration, marketing cheating, false transactions, false merchants, gambling, etc. There is no accurate manual marking or case characterization of such risk cases, and the black sample generator will not take the initiative to "report the case", lacking an effective risk feedback mechanism. As a result, during model training, the labels of some training samples are inaccurate, which affects subs...

AI Technical Summary

Problems solved by technology

There is no accurate manual marking or case characterization of such risk cases, and the black sample generator will not take the initiative to "report the case", lacking an effective risk feedback mechanism

As a result, during model training, the labels of some training samples are inaccurate, which affects subsequent model training

Images