Empirical Differential Privacy
arXiv:1910.12820v54 citations
Originality Incremental advance
AI Analysis
This addresses privacy concerns in data analysis by potentially improving utility while maintaining privacy, though it appears incremental as it builds on prior noiseless privacy work.
The paper tackles the problem of achieving differential privacy without or with reduced added noise by leveraging empirical noise in the data, avoiding explicit assumptions about the data-generating process.
We show how to achieve differential privacy with no or reduced added noise, based on the empirical noise in the data itself. Unlike previous works on noiseless privacy, the empirical viewpoint avoids making any explicit assumptions about the random process generating the data.