BriefGPT.xyz
Jan, 2019
机器学习的贝叶斯差分隐私
Improved Accounting for Differentially Private Learning
HTML
PDF
Aleksei Triastcyn, Boi Faltings
TL;DR
提出了基于贝叶斯的差分隐私方法，并实验表明基于样本数据分布的隐私保护方法比传统方法更实用有效。
Abstract
We consider the problem of
differential privacy
accounting, i.e. estimation of privacy loss bounds, in
machine learning
in a broad sense. We propose two versions of a generic privacy accountant suitable for a wid
→