R\'enyi Differential Privacy Mechanisms for Posterior Sampling
Machine Learning
2017-10-04 v1 Artificial Intelligence
Cryptography and Security
Abstract
Using a recently proposed privacy definition of R\'enyi Differential Privacy (RDP), we re-examine the inherent privacy of releasing a single sample from a posterior distribution. We exploit the impact of the prior distribution in mitigating the influence of individual data points. In particular, we focus on sampling from an exponential family and specific generalized linear models, such as logistic regression. We propose novel RDP mechanisms as well as offering a new RDP analysis for an existing method in order to add value to the RDP framework. Each method is capable of achieving arbitrary RDP privacy guarantees, and we offer experimental results of their efficacy.
Keywords
Cite
@article{arxiv.1710.00892,
title = {R\'enyi Differential Privacy Mechanisms for Posterior Sampling},
author = {Joseph Geumlek and Shuang Song and Kamalika Chaudhuri},
journal= {arXiv preprint arXiv:1710.00892},
year = {2017}
}
Comments
to be published in NIPS 2017