Method and device for authenticating user equipment according to misjudgment type

Abstract

The invention discloses a method and a device for authenticating user equipment according to misjudgment types. The method comprises the following steps of: configuring different loss weights for different misjudgment types in an equipment authentication model; selecting a specified equipment authentication model from the equipment authentication models according to the service scene; acquiring a loss function corresponding to a specified equipment authentication model based on the different loss weights; training the specified equipment authentication model by adopting a training sample, so that the loss function of the specified equipment authentication model in the training sample is minimum; and authenticating the new user equipment applying for the Internet service based on the trained specified equipment authentication model. According to the method, different loss weights are configured for different misjudgment types in the equipment authentication model, so that different misjudgment types can obtain different loss punishment, the authentication distinction degree and the authentication accuracy are improved, the risk or illegal equipment user authentication is prevented from passing, and the service safety is improved.

Description

technical field [0001] The present invention relates to the technical field of data processing, in particular to a method, device, electronic device and computer-readable medium for authenticating user equipment according to types of misjudgment. Background technique [0002] With the development of the Internet, various Internet service platforms have emerged, such as online shopping platforms, online car-hailing platforms, sharing platforms, maps, music and other Internet-based service platforms. When there are new user equipments applying for these Internet services, it is usually necessary to authenticate these new user equipments, so as to exclude those user equipments that have risks such as fraud and dishonesty. [0003] Currently, machine learning models are used to authenticate new user devices and classify them as "good" or "bad." However, the current machine learning model (such as: LR model, XGBoost model, etc.) defaults to the same loss caused by misjudgment of...

AI Technical Summary

Problems solved by technology

Since this authentication method gives the same penalty to all misjudgments, it is impossible to give different or more penalties to devices for misjudgment types, resulting in low authentication differentiation, poor authentication effects, and risky or illegal device users will be authenticated. If passed, there will be hidden dangers to platform business security

Images