Differentially Private Database Release via Kernel Mean Embeddings

TL;DR

This paper introduces a new method for privately releasing database summaries using kernel mean embeddings, enabling accurate population statistic estimation while protecting individual privacy.

Contribution

It proposes a novel framework that releases kernel mean embeddings for differentially private data sharing, with theoretical guarantees and two practical instantiations.

Findings

Guarantees differential privacy of the released embeddings.

Ensures the consistency of estimators derived from the embeddings.

Provides two implementations suitable for different scenarios.

Abstract

We lay theoretical foundations for new database release mechanisms that allow third-parties to construct consistent estimators of population statistics, while ensuring that the privacy of each individual contributing to the database is protected. The proposed framework rests on two main ideas. First, releasing (an estimate of) the kernel mean embedding of the data generating random variable instead of the database itself still allows third-parties to construct consistent estimators of a wide class of population statistics. Second, the algorithm can satisfy the definition of differential privacy by basing the released kernel mean embedding on entirely synthetic data points, while controlling accuracy through the metric available in a Reproducing Kernel Hilbert Space. We describe two instantiations of the proposed framework, suitable under different scenarios, and prove theoretical results guaranteeing differential privacy of the resulting algorithms and the consistency of estimators constructed from their outputs.