BriefGPT.xyz
Oct, 2024
精确的最小最大最优局部差分隐私采样
Exactly Minimax-Optimal Locally Differentially Private Sampling
HTML
PDF
Hyun-Young Park, Shahab Asoodeh, Si-Hyeon Lee
TL;DR
本研究针对局部差分隐私下的采样问题,填补了隐私-效用权衡的基础分析空白。我们定义了私有采样的最小最大效用-隐私权衡,并为有限和连续数据空间精确刻画了该权衡,提出了针对所有f-散度的普适最优采样机制。我们的实验表明,该机制在理论效用和经验效用方面均优于基线模型。
Abstract
The
Sampling
problem under local
Differential Privacy
has recently been studied with potential applications to generative models, but a fundamental analysis of its privacy-utility trade-off (PUT) remains incomple
→