On Choosing the $μ$ Parameter in Gaussian Differential Privacy
Researchers have published a paper detailing methods for converting privacy parameters between pure differential privacy ($\varepsilon$) and Gaussian differential privacy (GDP, $\mu$). The study proposes principled mappings by aligning worst-case membership inference attack success rates across three metrics. The authors recommend a general-purpose conversion of $\mu \approx \varepsilon/5$ for conservative privacy reporting in machine learning. AI
IMPACT Provides a standardized method for reporting privacy guarantees in machine learning models, potentially improving transparency and comparability.