English

Differentially Private Quasi-Concave Optimization: Bypassing the Lower Bound and Application to Geometric Problems

Cryptography and Security 2025-04-29 v1

Abstract

We study the sample complexity of differentially private optimization of quasi-concave functions. For a fixed input domain X\mathcal{X}, Cohen et al. (STOC 2023) proved that any generic private optimizer for low sensitive quasi-concave functions must have sample complexity Ω(2logX)\Omega(2^{\log^*|\mathcal{X}|}). We show that the lower bound can be bypassed for a series of ``natural'' problems. We define a new class of \emph{approximated} quasi-concave functions, and present a generic differentially private optimizer for approximated quasi-concave functions with sample complexity O~(logX)\tilde{O}(\log^*|\mathcal{X}|). As applications, we use our optimizer to privately select a center point of points in dd dimensions and \emph{probably approximately correct} (PAC) learn dd-dimensional halfspaces. In previous works, Bun et al. (FOCS 2015) proved a lower bound of Ω(logX)\Omega(\log^*|\mathcal{X}|) for both problems. Beimel et al. (COLT 2019) and Kaplan et al. (NeurIPS 2020) gave an upper bound of O~(d2.52logX)\tilde{O}(d^{2.5}\cdot 2^{\log^*|\mathcal{X}|}) for the two problems, respectively. We improve the dependency of the upper bounds on the cardinality of the domain by presenting a new upper bound of O~(d5.5logX)\tilde{O}(d^{5.5}\cdot\log^*|\mathcal{X}|) for both problems. To the best of our understanding, this is the first work to reduce the sample complexity dependency on X|\mathcal{X}| for these two problems from exponential in logX\log^* |\mathcal{X}| to logX\log^* |\mathcal{X}|.

Keywords

Cite

@article{arxiv.2504.19001,
  title  = {Differentially Private Quasi-Concave Optimization: Bypassing the Lower Bound and Application to Geometric Problems},
  author = {Kobbi Nissim and Eliad Tsfadia and Chao Yan},
  journal= {arXiv preprint arXiv:2504.19001},
  year   = {2025}
}
R2 v1 2026-06-28T23:12:31.250Z