Maximal α-Leakage and its Properties

摘要

Maximal α-leakage is a tunable measure of information leakage based on the quality of an adversary’s belief about an arbitrary function of private data based on public data. The parameter α determines the loss function used to measure the quality of a belief, ranging from log-loss at $lpha=1$ to the probability of error at $lpha=infty$. We review its definition and main properties, including extensions to $lphalt 1$, robustness to side information, and relationship to Rènyi differential privacy.