Instance-Optimal Differentially Private Estimation
Statistics Theory
2022-10-31 v1 Cryptography and Security
Machine Learning
Statistics Theory
Abstract
In this work, we study local minimax convergence estimation rates subject to -differential privacy. Unlike worst-case rates, which may be conservative, algorithms that are locally minimax optimal must adapt to easy instances of the problem. We construct locally minimax differentially private estimators for one-parameter exponential families and estimating the tail rate of a distribution. In these cases, we show that optimal algorithms for simple hypothesis testing, namely the recent optimal private testers of Canonne et al. (2019), directly inform the design of locally minimax estimation algorithms.
Cite
@article{arxiv.2210.15819,
title = {Instance-Optimal Differentially Private Estimation},
author = {Audra McMillan and Adam Smith and Jon Ullman},
journal= {arXiv preprint arXiv:2210.15819},
year = {2022}
}