Confidential-DPproof: Confidential Proof of Differentially Private Training - Inserm - Institut national de la santé et de la recherche médicale Accéder directement au contenu
Communication Dans Un Congrès Année : 2024

Confidential-DPproof: Confidential Proof of Differentially Private Training

Résumé

Post hoc privacy auditing techniques can be used to test the privacy guarantees of a model, but come with several limitations: (i) they can only establish lower bounds on the privacy loss, (ii) the intermediate model updates and some data must be shared with the auditor to get a better approximation of the privacy loss, and (iii) the auditor typically faces a steep computational cost to run a large number of attacks. In this paper, we propose to proactively generate a cryptographic certificate of privacy during training to forego such auditing limitations. We introduce Confidential-DPproof , a framework for Confidential Proof of Differentially Private Training, which enhances training with a certificate of the (ε, δ)-DP guarantee achieved. To obtain this certificate without revealing information about the training data or model, we design a customized zero-knowledge proof protocol tailored to the requirements introduced by differentially private training, including random noise addition and privacy amplification by subsampling. In experiments on CIFAR-10, Confidential-DPproof trains a model achieving state-of-the-art 91% test accuracy with a certified privacy guarantee of (ε = 0.55, δ = 10−5)-DP in approximately 100 hours.
Fichier principal
Vignette du fichier
8068_confidential_dpproof_confident.pdf (367.43 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04610635 , version 1 (13-06-2024)

Identifiants

  • HAL Id : hal-04610635 , version 1

Citer

Ali Shahin Shamsabadi, Gefei Tan, Tudor Ioan Cebere, Aurélien Bellet, Hamed Haddadi, et al.. Confidential-DPproof: Confidential Proof of Differentially Private Training. ICLR 2024 - 12th International Conference on Learning Representations, May 2024, Vienna (Austria), Austria. ⟨hal-04610635⟩
8 Consultations
6 Téléchargements

Partager

Gmail Mastodon Facebook X LinkedIn More