Concurrent Composition for Interactive Differential Privacy with Adaptive Privacy-Loss Parameters
Open Access
- 15 November 2023
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1949-1963
Abstract
In this paper, we study the concurrent composition of interactive mechanisms with adaptively chosen privacy-loss parameters. In this setting, the adversary can interleave queries to existing interactive mechanisms, as well as create new ones. We prove that every valid privacy filter and odometer for noninteractive mechanisms extends to the concurrent composition of interactive mechanisms if privacy loss is measured using (ε, δ)-DP, ƒ-DP, or Rényi DP of fixed order. Our results offer strong theoretical foundations for enabling full adaptivity in composing differentially private interactive mechanisms, showing that concurrency does not affect the privacy guarantees. We also provide an implementation for users to deploy in practice.Keywords
All Related Versions
Funding Information
- National Science Foundation (BCS-2218803)
- Simons Foundation
- Alfred P. Sloan Foundation
- Apple
- U.S. Census Bureau
- Computing Research Association
This publication has 6 references indexed in Scilit:
- Rényi Differential Privacy2017
- Concentrated Differential Privacy: Simplifications, Extensions, and Lower Bounds2016
- The Complexity of Computing the Optimal Composition of Differential Privacy2015
- Rényi Divergence and Kullback-Leibler DivergenceIEEE Transactions on Information Theory, 2014
- Boosting and Differential Privacy2010
- Our Data, Ourselves: Privacy Via Distributed Noise Generation2006