Research map / Learning theory and algorithms / Kernels and regression theory
Parameter-free HE-friendly Logistic Regression
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| replace LRHE (sigmoid approximation + gradient descent) with ridge regression on teacher logits | Ours · Adult | accuracy (%): 81.913 -> 83.123 | Helped |
| replace LRHE with ridge regression on teacher logits | Ours · Diabetes | accuracy (%): 71.429 -> 75.325 | Helped |
| replace LRHE with closed-form ridge regression with partial encryption | Ours · Adult | computation time (sec): 306.061 -> 102.25 | Helped |
| add mean matching of teacher logits (ablation) | Ours · Diabetes (shifted split) | accuracy (%): 29.487 -> 80.769 | Helped |
| add mean matching of teacher logits (ablation) | Ours · Adult (shifted split) | accuracy (%): 63.330 -> 69.471 | Helped |
| add mean matching of teacher logits (ablation) | Ours · Credit (shifted split) | accuracy (%): 20.833 -> 79.167 | Helped |
| train step 3 with 8 iterations of encrypted gradient descent instead of closed form (ablation) | Ours · Diabetes | accuracy (%): 75.325 -> 76.623 | Helped |
| train step 3 with gradient descent instead of closed form (ablation) | Ours · Adult | computation time (sec): 102.25 -> 1488.65 | Hurt |
| train step 3 with gradient descent instead of closed form (ablation) | Ours · Cancer | accuracy (%): 90.580 -> 90.580 | No clear effect |
Closest papers by meaning
- Certification for Differentially Private Prediction in Gradient-Based Training (ICML 2025 · similarity 0.85)
- Privacy-Preserving Embedding via Look-up Table Evaluation with Fully Homomorphic Encryption (ICML 2024 · similarity 0.85)
- Revisiting Differentially Private ReLU Regression (NeurIPS 2024 · similarity 0.85)
- Online Differentially Private Conformal Prediction for Uncertainty Quantification (ICML 2025 · similarity 0.85)
- Differentially Private Model Merging (NeurIPS 2026 · similarity 0.84)
- Online Sensitivity Optimization in Differentially Private Learning (AAAI 2024 · similarity 0.84)
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.