Data set authentication method and system based on machine learning member inference attack

Abstract

The invention discloses a data set authentication method and system based on a machine learning member inference attack, and belongs to the field of data protection of the Internet of Things, and the method comprises the steps: selecting a plurality of machine learning models after obtaining a target data set and an auxiliary data set, and respectively constructing reference model groups based on the two data sets; predicting the target data set by using the two types of reference model groups to obtain a member prediction set and a non-member prediction set; taking the member prediction set and the non-member prediction set as features, taking corresponding member attributes as labels, and training to obtain an authentication model; performing member inference attack on all data in the member prediction set by using the authentication model, and screening member fingerprint data from the target data set; and based on the authentication model, obtaining the probability that the member fingerprint data is member data of the suspicious model, thereby determining whether the suspicious model is obtained by training an Interne-of-Things data set. Therefore, the interests and privacy of the data owner can be effectively protected.

Description

technical field [0001] The invention belongs to the field of Internet of Things data protection, and more specifically relates to a data set authentication method and system based on machine learning membership inference attack. Background technique [0002] The explosion of IoT data in recent years has fueled the application of machine learning in everything from natural language processing to computer vision. However, obtaining data from raw data that can be used to train machine learning models is a non-trivial task, especially at an industrial level. Specifically, collecting data requires a lot of cost and expertise, including data collection, data annotation, and data preprocessing. Furthermore, from a privacy point of view, the data may contain some sensitive records related to the data provider, and the owner of the data set is responsible for the protection of the data held by himself as mentioned in the General Data Protection Regulation; moreover , data itself is...

AI Technical Summary

Problems solved by technology

At present, the protection of data sets is mainly reflected in the field of data publishing. It mainly uses technologies such as differential privacy to blur private data. This does protect some private data to a large extent, but it is impossible to verify whether the data set is used for training. specific model

In practice, machine learning models are often deployed as black boxes, where access to suspect models is limited to black-box interfaces that output model predictions on given Indirect relationships are also difficult to verify

Images