Empirical Differential Privacy

TL;DR

This paper introduces a method for achieving differential privacy by leveraging the empirical noise in data, reducing or eliminating the need for added noise, and avoiding assumptions about data generation processes.

Contribution

It presents a novel empirical approach to differential privacy that minimizes added noise without relying on traditional probabilistic assumptions.

Findings

Achieves differential privacy with reduced or no added noise.

Avoids assumptions about the data's random process.

Provides a practical framework for empirical privacy preservation.

Abstract

We show how to achieve differential privacy with no or reduced added noise, based on the empirical noise in the data itself. Unlike previous works on noiseless privacy, the empirical viewpoint avoids making any explicit assumptions about the random process generating the data.