Risk assessment model training, risk assessment method and device

Abstract

The embodiments of this specification provide a training of a risk assessment model, a risk assessment method, and an apparatus. In the training method, a batch of user samples are collected, and each user sample includes user characteristics and a risk label, and the risk label is used to indicate the user , as well as multiple individual risk scores for each predetermined risk type. Based on the user characteristics of each user sample and the total risk score in the risk label, the feature coding layer is trained to obtain a trained feature coding layer, and based on the trained feature coding layer, the feature coding result of each user sample is obtained. Based on the feature encoding results of each user sample and the total risk score in the risk label, the comprehensive risk assessment layer is trained. For each single risk assessment layer in each single risk assessment layer, the single risk assessment layer is trained based on the feature coding result of each user sample and the individual risk score of the corresponding risk type in the risk label.

Description

technical field [0001] One or more embodiments of the present specification relate to the technical field of machine learning, and in particular, to a risk assessment model training, a risk assessment method, and an apparatus. Background technique [0002] With the increasing popularity of network technology, there are more and more risks in network transactions. In order to effectively reduce high-risk transactions and strengthen market monitoring, it is extremely important to conduct risk assessment on users. [0003] In traditional technologies, risk assessment is usually performed only for a single risk type (eg, a risk type suspected of gambling). The specific evaluation process may be as follows: first, based on the sample users and normal sample users of the risk type, the risk evaluation model is trained. Then, based on the risk assessment model, assess whether the target user is a user of the corresponding risk type. Therefore, traditional risk assessment methods...

AI Technical Summary

Problems solved by technology

Therefore, traditional risk assessment methods are less efficient

Images