English

Computing Differential Privacy Guarantees for Heterogeneous Compositions Using FFT

Cryptography and Security 2021-06-22 v2 Machine Learning Machine Learning

Abstract

The recently proposed Fast Fourier Transform (FFT)-based accountant for evaluating (ε,δ)(\varepsilon,\delta)-differential privacy guarantees using the privacy loss distribution formalism has been shown to give tighter bounds than commonly used methods such as R\'enyi accountants when applied to homogeneous compositions, i.e., to compositions of identical mechanisms. In this paper, we extend this approach to heterogeneous compositions. We carry out a full error analysis that allows choosing the parameters of the algorithm such that a desired accuracy is obtained. The analysis also extends previous results by taking into account all the parameters of the algorithm. Using the error analysis, we also give a bound for the computational complexity in terms of the error which is analogous to and slightly tightens the one given by Murtagh and Vadhan (2018). We also show how to speed up the evaluation of tight privacy guarantees using the Plancherel theorem at the cost of increased pre-computation and memory usage.

Keywords

Cite

@article{arxiv.2102.12412,
  title  = {Computing Differential Privacy Guarantees for Heterogeneous Compositions Using FFT},
  author = {Antti Koskela and Antti Honkela},
  journal= {arXiv preprint arXiv:2102.12412},
  year   = {2021}
}

Comments

44 pages, 10 figures

R2 v1 2026-06-23T23:28:49.649Z