d3p - A Python Package for Differentially-Private Probabilistic Programming

Loading...
Thumbnail Image

Access rights

openAccess
publishedVersion

URL

Journal Title

Journal ISSN

Volume Title

A4 Artikkeli konferenssijulkaisussa

Date

2022-04-01

Major/Subject

Mcode

Degree programme

Language

en

Pages

Series

Proceedings of Privacy Enhancing Technologies, Volume 2022, issue 2, pp. 407–425

Abstract

We present d3p, a software package designed to help fielding runtime efficient widely-applicable Bayesian inference under differential privacy guarantees. d3p achieves general applicability to a wide range of probabilistic modelling problems by implementing the differentially private variational inference algorithm, allowing users to fit any parametric probabilistic model with a differentiable density function. d3p adopts the probabilistic programming paradigm as a powerful way for the user to flexibly define such models. We demonstrate the use of our software on a hierarchical logistic regression example, showing the expressiveness of the modelling approach as well as the ease of running the parameter inference. We also perform an empirical evaluation of the runtime of the private inference on a complex model and find an ~10 fold speed-up compared to an implementation using TensorFlow Privacy.

Description

Keywords

Differential privacy, JAX, NumPyro, Probabilistic programming, Variational inference

Other note

Citation

Prediger, L, Loppi, N, Kaski, S & Honkela, A 2022, ' d3p - A Python Package for Differentially-Private Probabilistic Programming ', Proceedings of Privacy Enhancing Technologies, vol. 2022, no. 2, pp. 407–425 . https://doi.org/10.2478/popets-2022-0052