BriefGPT.xyz
May, 2020
差分隐私中的近似最优性
Near Instance-Optimality in Differential Privacy
HTML
PDF
Hilal Asi, John C. Duchi
TL;DR
本文介绍了差分隐私中的两种实例优化概念，并提出了一种补充试点机制，称为反敏感机制，可针对一类估计量进行实例优化。此外，这些机制在多个函数类别的每个实例上都可以优于平滑敏感性框架。
Abstract
We develop two notions of instance optimality in
differential privacy
, inspired by classical statistical theory: one by defining a
local minimax risk
and the other by considering
→