I have been studying Differential Privacy (DP) and there’s something I do not understand: why is the variance of the noise in most DP schemes constant across all individuals (assuming a one-to-one relationship between individuals and data points). Some individuals may be close to the statistic that DP allows to compute to privately and thus may require less noise than outliers in the dataset. As a result, DP is over-protecting those individuals. Wouldn’t it be more efficient (greater utility) to adapt the noise according to the information that each individual data point reveals?
Information Security Stack Exchange is a question and answer site for information security professionals. It only takes a minute to sign up.
Sign up to join this community
Anybody can ask a question
Anybody can answer
The best answers are voted up and rise to the top
