α-leakage by Rényi Divergence and Sibson Mutual Information

2024/05/01 by Ding, Ni, Zarrabian, Mohammad Amin, Sadeghi, Parastoo · 1 citation
#FOS: Computer and information sciences #Information Theory (cs.IT)

paper · doi:10.48550/arxiv.2405.00423

Abstract

For f(t) = exp(\fracα-1αt), this paper proposes a f-mean information gain measure. Rényi divergence is shown to be the maximum f-mean information gain incurred at each elementary event y of channel output Y and Sibson mutual information is the f-mean of this Y-elementary information gain. Both are proposed as α-leakage measures, indicating the most information an adversary can obtain on sensitive data. It is shown that the existing α-leakage by Arimoto mutual information can be expressed as f-mean measures by a scaled probability. Further, Sibson mutual information is interpreted as the maximum f-mean information gain over all estimation decisions applied to channel output.

Cited by

• α-leakage Interpretation of Rényi Capacity

Related

• α-leakage Interpretation of Rényi Capacity
• A Tunable Measure for Information Leakage
• Robustness of Maximal α-Leakage to Side Information
• An Operational Approach to Information Leakage via Generalized Gain Functions
• Robust Generalization via α-Mutual Information