Rapid model forgetting method and system based on generative adversarial network

Abstract

The invention discloses a rapid model forgetting method and system based on a generative adversarial network, and the method comprises the steps: inputting third-party data with the same distribution as to-be-forgotten data into an original model, sorting the output results of the original model, obtaining a first sorting result, initializing a generator into the original model, inputting the to-be-forgotten data into the generator, and obtaining a second sorting result; and sorting results output by the generator to obtain a second sorting result, alternately training the generator and the discriminator by using the first sorting result and the second sorting result, and stopping training until the discriminator cannot distinguish the distribution difference between the output of the to-be-forgotten data on the generator and the output of the third-party data on the original model. Member reasoning attacks are carried out on the generator, if an attack result is to-be-forgotten data and is not trained by the generator, forgetting succeeds, and the trained generator is used as a forgotten model; according to the method, the speed of forgetting the data in the model can be increased, and the effect is more obvious especially in a complex scene.

Description

technical field [0001] The invention relates to the technical field of deep learning, in particular to a method and system for fast model forgetting based on a generative confrontation network. Background technique [0002] In machine learning, especially in over-parameterized deep learning, the model trained on data will memorize a lot of information about the training data, which will bring serious privacy and security problems to users, such as attackers getting the model After the model is reversed, the data for training the model can be recovered from the trained model. Even if the attacker only queries the output of the model without knowing the internal details of the model, the member inference attack can also determine a certain data. whether to be used to train the model. The existence of these attacks shows that the model's memory of the training data will pose a serious threat to the user's privacy. Therefore, in order to protect the privacy of users, some laws...

AI Technical Summary

Problems solved by technology

[0004] The embodiment of the present application provides a fast model forgetting method and system based on generative confrontation

Images