The Edgeworth Accountant uses the Edgeworth expansion on privacy-loss log-likelihood ratios to derive closed-form non-asymptotic (ε, δ)-DP bounds for composed noise-addition mechanisms.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2022 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Edgeworth Accountant: An Analytical Approach to Differential Privacy Composition
The Edgeworth Accountant uses the Edgeworth expansion on privacy-loss log-likelihood ratios to derive closed-form non-asymptotic (ε, δ)-DP bounds for composed noise-addition mechanisms.