Differentially Private Quantiles with Smaller Error

TL;DR

This paper introduces a new differentially private algorithm for approximate quantiles that achieves smaller error bounds by using continual counting techniques, especially effective when the number of quantiles is large.

Contribution

It presents a novel differentially private quantile mechanism with improved error bounds and a correlated randomization approach, also extending to $( ext{ε}, ext{δ})$-DP without the gap assumption.

Findings

Achieves lower maximum rank error compared to prior algorithms.

Performs favorably in experiments, especially with many quantiles.

Extends to $( ext{ε}, ext{δ})$-DP, relaxing previous assumptions.

Abstract

In the approximate quantiles problem, the goal is to output $m$ quantile estimates, the ranks of which are as close as possible to $m$ given quantiles $0 \leq q_1 \leq\dots \leq q_m \leq 1$. We present a mechanism for approximate quantiles that satisfies $\varepsilon$-differential privacy for a dataset of $n$ real numbers where the ratio between the distance between the closest pair of points and the size of the domain is bounded by $\psi$. As long as the minimum gap between quantiles is sufficiently large, $|q_i-q_{i-1}|\geq \Omega\left(\frac{m\log(m)\log(\psi)}{n\varepsilon}\right)$ for all $i$, the maximum rank error of our mechanism is $O\left(\frac{\log(\psi) + \log^2(m)}{\varepsilon}\right)$ with high probability. Previously, the best known algorithm under pure DP was due to Kaplan, Schnapp, and Stemmer~(ICML '22), who achieved a bound of $O\left(\frac{\log(\psi)\log^2(m) + \log^3(m)}{\varepsilon}\right)$. Our improvement stems from the use of continual counting techniques which allows the quantiles to be randomized in a correlated manner. We also present an $(\varepsilon,\delta)$-differentially private mechanism that relaxes the gap assumption without affecting the error bound, improving on existing methods when $\delta$ is sufficiently close to zero. We provide experimental evaluation which confirms that our mechanism performs favorably compared to prior work in practice, in particular when the number of quantiles $m$ is large.