Research map / Federated learning and privacy / Differential privacy
Understanding Private Learning From Feature Perspective
Meng Ding, Mingxi Lei, Shaopeng Fu, Shaowei Wang, Di Wang, Jinhui Xu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use a moderate privacy budget (epsilon 1 or 5) at SNR 0.6 (ablation) | NoisyGD two-layer CNN · synthetic multi-patch data | feature signal vs noise memorization | No clear effect |
| lower privacy budget to epsilon 0.2 at SNR 0.6 (ablation) | NoisyGD two-layer CNN · synthetic multi-patch data | feature signal vs noise memorization | Hurt |
| increase SNR from 0.6 to 3 at epsilon 0.2 (ablation) | NoisyGD two-layer CNN · synthetic multi-patch data | feature signal learning | Improved |
| apply noisy gradient descent at SNR 0.2 where non-private training memorizes noise (ablation) | NoisyGD two-layer CNN · synthetic multi-patch data | data noise memorization | Hurt |
| reduce privacy budget epsilon at low SNR (0.2) (ablation) | NoisyGD two-layer CNN · synthetic multi-patch data | noise memorization of other filters | Hurt |
Closest papers by meaning
- The importance of feature preprocessing for differentially private linear optimization (ICLR 2024 · similarity 0.89)
- Differentially Private Prototypes for Imbalanced Transfer Learning (AAAI 2025 · similarity 0.88)
- Differentially Private Stochastic Gradient Descent with Fixed-Size Minibatches: Tighter RDP Guarantees with or without Replacement (NeurIPS 2024 · similarity 0.88)
- DOPPLER: Differentially Private Optimizers with Low-pass Filter for Privacy Noise Reduction (NeurIPS 2024 · similarity 0.87)
- Differentially private optimization for non-decomposable objective functions (ICLR 2025 · similarity 0.87)
- Subsampling is not Magic: Why Large Batch Sizes Work for Differentially Private Stochastic Optimisation (ICML 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
