Subsampled Rényi Differential Privacy and Analytical Moments Accountant

07/31/2018
by   Yu-Xiang Wang, et al.
2

We study the problem of subsampling in differential privacy (DP), a question that is the centerpiece behind many successful differentially private machine learning algorithms. Specifically, we provide a tight upper bound on the Rényi Differential Privacy (RDP) (Mironov, 2017) parameters for algorithms that: (1) subsample the dataset, and then (2) apply a randomized mechanism M to the subsample, in terms of the RDP parameters of M and the subsampling probability parameter. This result generalizes the classic subsampling-based "privacy amplification" property of (ϵ,δ)-differential privacy that applies to only one fixed pair of (ϵ,δ), to a stronger version that exploits properties of each specific randomized algorithm and satisfies an entire family of (ϵ(δ),δ)-differential privacy for all δ∈ [0,1]. Our experiments confirm the advantage of using our techniques over keeping track of (ϵ,δ) directly, especially in the setting where we need to compose many rounds of data access.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
01/01/2022

Differential Privacy Made Easy

Data privacy is a major issue for many decades, several techniques have ...
research
10/14/2017

Learners that Leak Little Information

We study learning algorithms that are restricted to revealing little inf...
research
03/04/2021

Quantifying identifiability to choose and audit ε in differentially private deep learning

Differential privacy allows bounding the influence that training data re...
research
11/19/2022

A Survey on Differential Privacy with Machine Learning and Future Outlook

Nowadays, machine learning models and applications have become increasin...
research
05/22/2023

Differential Privacy with Random Projections and Sign Random Projections

In this paper, we develop a series of differential privacy (DP) algorith...
research
04/01/2020

Differential Privacy for Sequential Algorithms

We study the differential privacy of sequential statistical inference an...
research
02/23/2015

Learning with Differential Privacy: Stability, Learnability and the Sufficiency and Necessity of ERM Principle

While machine learning has proven to be a powerful data-driven solution ...

Please sign up or login with your details

Forgot password? Click here to reset