Correlation inference attacks against machine learning models
File(s)
Published version
Author(s)
Cretu, Ana maria
Guepin, Florent
de Montjoye, Yves Alexandre
Type
Journal Article
Abstract
Despite machine learning models being widely used today, the relationship between a model and its training dataset is not well understood. We explore correlation inference attacks, whether and when a model leaks information about the correlations between the input variables of its training dataset. We first propose a model-less attack, where an adversary exploits the spherical parametrization of correlation matrices alone to make an informed guess. Second, we propose a model-based attack, where an adversary exploits black-box model access to infer the correlations using minimal and realistic assumptions.
Third, we evaluate our attacks against logistic regression and multilayer perceptron models on three tabular datasets and show the models to leak correlations. We finally show how extracted correlations can be used as building blocks for attribute inference attacks and enable weaker adversaries. Our results raise fundamental questions on what a model does and should remember from its training set.
Third, we evaluate our attacks against logistic regression and multilayer perceptron models on three tabular datasets and show the models to leak correlations. We finally show how extracted correlations can be used as building blocks for attribute inference attacks and enable weaker adversaries. Our results raise fundamental questions on what a model does and should remember from its training set.
Date Issued
2024-07
Date Acceptance
2024-06-03
Citation
Science Advances, 2024, 10 (28)
ISSN
2375-2548
Publisher
American Association for the Advancement of Science
Journal / Book Title
Science Advances
Volume
10
Issue
28
Copyright Statement
Copyright © 2024 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. Distributed under a Creative Commons Attribution License 4.0 (CC BY).
https://creativecommons.org/licenses/by/4.0/
This is an open-access article distributed under the terms of the Creative Commons Attribution license, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
https://creativecommons.org/licenses/by/4.0/
This is an open-access article distributed under the terms of the Creative Commons Attribution license, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
License URL
Identifier
https://www.science.org/doi/10.1126/sciadv.adj9260
Publication Status
Published
Article Number
eadj9260
Date Publish Online
2024-07-10