Differentially Private Normalizing Flows for Privacy-Preserving Density Estimation

TL;DR

This paper introduces a differentially private normalizing flow model that enables privacy-preserving density estimation and anomaly detection, outperforming previous methods on benchmark datasets.

Contribution

We propose a novel normalizing flow approach that guarantees differential privacy for density estimation, addressing privacy concerns in sensitive data applications.

Findings

Outperforms previous state-of-the-art privacy-preserving density estimation methods.

Provides explicit differential privacy guarantees for normalizing flows.

Effective in differentially private anomaly detection tasks.

Abstract

Normalizing flow models have risen as a popular solution to the problem of density estimation, enabling high-quality synthetic data generation as well as exact probability density evaluation. However, in contexts where individuals are directly associated with the training data, releasing such a model raises privacy concerns. In this work, we propose the use of normalizing flow models that provide explicit differential privacy guarantees as a novel approach to the problem of privacy-preserving density estimation. We evaluate the efficacy of our approach empirically using benchmark datasets, and we demonstrate that our method substantially outperforms previous state-of-the-art approaches. We additionally show how our algorithm can be applied to the task of differentially private anomaly detection.