A Unified Learn-to-Distort-Data Framework for Privacy-Utility Trade-off in Trustworthy Federated Learning

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Zhang, Xiaojin, Xu, Mingcong, Chen, Wei
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909247939280896
author Zhang, Xiaojin
Xu, Mingcong
Chen, Wei
author_facet Zhang, Xiaojin
Xu, Mingcong
Chen, Wei
contents In this paper, we first give an introduction to the theoretical basis of the privacy-utility equilibrium in federated learning based on Bayesian privacy definitions and total variation distance privacy definitions. We then present the \textit{Learn-to-Distort-Data} framework, which provides a principled approach to navigate the privacy-utility equilibrium by explicitly modeling the distortion introduced by the privacy-preserving mechanism as a learnable variable and optimizing it jointly with the model parameters. We demonstrate the applicability of our framework to a variety of privacy-preserving mechanisms on the basis of data distortion and highlight its connections to related areas such as adversarial training, input robustness, and unlearnable examples. These connections enable leveraging techniques from these areas to design effective algorithms for privacy-utility equilibrium in federated learning under the \textit{Learn-to-Distort-Data} framework.
format Preprint
id arxiv_https___arxiv_org_abs_2407_04751
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Unified Learn-to-Distort-Data Framework for Privacy-Utility Trade-off in Trustworthy Federated Learning
Zhang, Xiaojin
Xu, Mingcong
Chen, Wei
Machine Learning
Artificial Intelligence
Cryptography and Security
In this paper, we first give an introduction to the theoretical basis of the privacy-utility equilibrium in federated learning based on Bayesian privacy definitions and total variation distance privacy definitions. We then present the \textit{Learn-to-Distort-Data} framework, which provides a principled approach to navigate the privacy-utility equilibrium by explicitly modeling the distortion introduced by the privacy-preserving mechanism as a learnable variable and optimizing it jointly with the model parameters. We demonstrate the applicability of our framework to a variety of privacy-preserving mechanisms on the basis of data distortion and highlight its connections to related areas such as adversarial training, input robustness, and unlearnable examples. These connections enable leveraging techniques from these areas to design effective algorithms for privacy-utility equilibrium in federated learning under the \textit{Learn-to-Distort-Data} framework.
title A Unified Learn-to-Distort-Data Framework for Privacy-Utility Trade-off in Trustworthy Federated Learning
topic Machine Learning
Artificial Intelligence
Cryptography and Security
url https://arxiv.org/abs/2407.04751