Private Mean Estimation of Heavy-Tailed Distributions
Gautam Kamath Thanks: Cheriton School of Computer Science, University of Waterloo. Vikrant Singhal Thanks: Khoury College of Computer Sciences, Northeastern University. Jonathan Ullman Thanks: Khoury College of Computer Sciences, Northeastern University.
Abstract
We give new upper and lower bounds on the minimax sample complexity of differentially private mean estimation of distributions with bounded $k$ -th moments. Roughly speaking, in the univariate case, we show that
原文 arXiv:2002.09464;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2002.09464v3