Derives a relative disclosure risk indicator (RDR) and algorithms for selecting epsilon in differential privacy based on within-dataset individual risks, plus a multi-query leakage bound.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
verdicts
UNVERDICTED 2representative citing papers
RESFL integrates adversarial feature disentanglement and uncertainty-aware client weighting in federated learning to reduce membership inference attacks and equality-of-opportunity gaps while preserving model utility for object detection.
citing papers explorer
-
Within-Dataset Disclosure Risk for Differential Privacy
Derives a relative disclosure risk indicator (RDR) and algorithms for selecting epsilon in differential privacy based on within-dataset individual risks, plus a multi-query leakage bound.
-
RESFL: An Uncertainty-Aware Framework for Responsible Federated Learning by Balancing Privacy, Fairness and Utility
RESFL integrates adversarial feature disentanglement and uncertainty-aware client weighting in federated learning to reduce membership inference attacks and equality-of-opportunity gaps while preserving model utility for object detection.