QOP achieves (ε, δ)-differential privacy for ERM in the interpolation regime under weaker assumptions than linear objective perturbation by using random quadratic curvature to enforce stability and control sensitivity.
Title resolution pending
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3verdicts
UNVERDICTED 3representative citing papers
FiBeR adds a closed-form filter-aware correction A(ω)σ_w² to the second-moment term for temporally filtered DP gradients, improving adaptive optimization performance.
CA-ADP adjusts differential privacy noise per mini-batch class composition to improve F-scores by 3.3-8.5% over standard DP on three fall-detection datasets while claiming formal (ε,δ) guarantees.
citing papers explorer
-
Quadratic Objective Perturbation: Curvature-Based Differential Privacy
QOP achieves (ε, δ)-differential privacy for ERM in the interpolation regime under weaker assumptions than linear objective perturbation by using random quadratic curvature to enforce stability and control sensitivity.
-
FIBER: A Differentially Private Optimizer with Filter-Aware Innovation Bias Correction
FiBeR adds a closed-form filter-aware correction A(ω)σ_w² to the second-moment term for temporally filtered DP gradients, improving adaptive optimization performance.
-
Class-Aware Adaptive Differential Privacy in Deep Learning for Sensor-Based Fall Detection
CA-ADP adjusts differential privacy noise per mini-batch class composition to improve F-scores by 3.3-8.5% over standard DP on three fall-detection datasets while claiming formal (ε,δ) guarantees.