核心概念
강화를 통해 그룹 프라이버시를 강화하고, 샘플링을 통한 통합된 증폭을 통해 Rényi Differential Privacy를 다룹니다.
統計
M = B ◦ S는 O(rϵ)-DP이다.
M = B ◦ S는 2O(rϵ)-DP이다.
Ψα(mx||mx′) ≤ Z Y2 cα(y(1), y(2)) dΓ((y(1), y(2)))
引用
"Our goal is to determine whether stronger privacy guarantees can be obtained by considering multiple of these properties jointly."
"These results establish the joint study of different DP properties as a promising research direction."