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.
The Conjugate Gradient Algorithm on Well-Conditioned
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
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.