Effects of Differential Privacy and Data Skewness on Membership Inference Vulnerability

Membership inference attacks seek to infer the membership of individual training instances of a privately trained model. This paper presents a membership privacy analysis and evaluation system, called MPLens, with three unique contributions... (read more)

Results in Papers With Code
(↓ scroll down to see all results)