English

Optimal Differentially Private Sampling of Unbounded Gaussians

Data Structures and Algorithms 2025-03-04 v1 Cryptography and Security Information Theory Machine Learning math.IT Machine Learning

Abstract

We provide the first O~(d)\widetilde{\mathcal{O}}\left(d\right)-sample algorithm for sampling from unbounded Gaussian distributions under the constraint of (ε,δ)\left(\varepsilon, \delta\right)-differential privacy. This is a quadratic improvement over previous results for the same problem, settling an open question of Ghazi, Hu, Kumar, and Manurangsi.

Keywords

Cite

@article{arxiv.2503.01766,
  title  = {Optimal Differentially Private Sampling of Unbounded Gaussians},
  author = {Valentio Iverson and Gautam Kamath and Argyris Mouzakis},
  journal= {arXiv preprint arXiv:2503.01766},
  year   = {2025}
}

Comments

47 pages

R2 v1 2026-06-28T22:04:59.519Z