Titre :

Differential Privacy for Bayesian Inference through Posterior Sampling

Auteur(s) :

Dimitrakakis, Christos [Auteur]
Chalmers University of Technology [Göteborg]
Sequential Learning [SEQUEL]
Université de Lille, Sciences Humaines et Sociales
Harvard University
Nelson, Blaine [Auteur]
Google Inc [Mountain View]
Zhang, Zuhe [Auteur]
University of Melbourne
Mitrokotsa, Aikateirni [Auteur]
Chalmers University of Technology [Göteborg]
Rubinstein, Benjamin [Auteur]
University of Melbourne

Titre de la revue :

Journal of Machine Learning Research

Pagination :

1−39

Éditeur :

Microtome Publishing

Date de publication :

2017-04

ISSN :

1532-4435

Mot(s)-clé(s) en anglais :

Bayesian inference
differential privacy
robustness
adversarial Learning

Discipline(s) HAL :

Statistiques [stat]/Machine Learning [stat.ML]
Informatique [cs]/Cryptographie et sécurité [cs.CR]
Mathématiques [math]/Statistiques [math.ST]

Résumé en anglais : [en]

Differential privacy formalises privacy-preserving mechanisms that provide access to a database. Can Bayesian inference be used directly to provide private access to data? The answer is yes: under certain conditions on the ...
Lire la suite >Differential privacy formalises privacy-preserving mechanisms that provide access to a database. Can Bayesian inference be used directly to provide private access to data? The answer is yes: under certain conditions on the prior, sampling from the posterior distribution can lead to a desired level of privacy and utility. For a uniform treatment, we define differential privacy over arbitrary data set metrics, outcome spaces and distribution families. This allows us to also deal with non-i.i.d or non-tabular data sets. We then prove bounds on the sensitivity of the posterior to the data, which delivers a measure of robustness. We also show how to use posterior sampling to provide differentially private responses to queries, within a decision-theoretic framework. Finally, we provide bounds on the utility of answers to queries and on the ability of an adversary to distinguish between data sets. The latter are complemented by a novel use of Le Cam's method to obtain lower bounds on distinguishability. Our results hold for arbitrary metrics, including those for the common definition of differential privacy. For specific choices of the metric, we give a number of examples satisfying our assumptions. *. A preliminary version of this paper appeared in Algorithmic Learning Theory 2014 (Dimitrakakis et al., 2014). This version corrects proofs, constant factors in the upper bounds and introduces new material on utility analysis, lower bounds and examples.Lire moins >

Langue :

Anglais

Comité de lecture :

Oui

Audience :

Internationale

Vulgarisation :

Non

Collections :

• Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL) - UMR 9189

Source :

Harvested from HAL