Jump to content

Unbounded differential privacy

From Hackerpedia
Revision as of 02:53, 15 January 2026 by imported>Unknown user

Languages: English | Français

Unbounded differential privacy

A unit of privacy variant that calls two datasets D1 and D2 neighbors if it is possible to construct D2 from D1 by adding or removing one person’s data. Under unbounded differential privacy, neighboring datasets have different sizes.


Source: NIST SP 800-226 | Category: