Skip to Main content Skip to Navigation
Conference papers

Privacy Assessment of Federated Learning using Private Personalized Layers

Théo Jourdan 1 Antoine Boutet 1 Carole Frindel 2 
1 PRIVATICS - Privacy Models, Architectures and Tools for the Information Society
Inria Grenoble - Rhône-Alpes, CITI - CITI Centre of Innovation in Telecommunications and Integration of services, Inria Lyon
Abstract : Federated Learning (FL) is a collaborative scheme to train a learning model across multiple participants without sharing data. While FL is a clear step forward towards enforcing usersâĂŹ privacy, different inference attacks have been developed. In this paper, we quantify the utility and privacy trade-off of a FL scheme using private personalized layers. While this scheme has been proposed as local adaptation to improve the accuracy of the model through local personalization, it has also the advantage to minimize the information about the model exchanged with the server. However, the privacy of such a scheme has never been quantified. Our evaluations using motion sensor dataset show that personalized layers speed up the convergence of the model and slightly improve the accuracy for all users compared to a standard FL scheme while better preventing both attribute and membership inferences compared to a FL scheme using local differential privacy.
Complete list of metadata
Contributor : Antoine Boutet Connect in order to contact the contributor
Submitted on : Sunday, September 26, 2021 - 5:19:43 AM
Last modification on : Saturday, September 24, 2022 - 2:44:04 PM
Long-term archiving on: : Monday, December 27, 2021 - 6:04:08 PM


Files produced by the author(s)



Théo Jourdan, Antoine Boutet, Carole Frindel. Privacy Assessment of Federated Learning using Private Personalized Layers. MLSP 2021 - IEEE International Workshop on Machine Learning for Signal Processing, Oct 2021, Queensland, Australia. pp.1-5, ⟨10.1109/MLSP52302.2021.9596237⟩. ⟨hal-03354722⟩



Record views


Files downloads