Research map / Federated learning and privacy / Differential privacy
Differentially Private Fair Learning
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 |
|---|---|---|---|
| Use private post-processing rather than private in-processing | Fair learning on Communities and Crime · Communities and Crime | error-fairness tradeoff | Helped |
| Use private in-processing rather than private post-processing | Fair learning on Communities and Crime · Communities and Crime | error-fairness tradeoff | Hurt |
| Allow protected group membership at test time | DP fair learning · Equalized false positive rate | error-fairness tradeoff | Helped |
| Use maximally discriminatory group-indicator classifiers in the hypothesis class (ablation) | DP-oracle-learner · Equalized false positive rate | sensitivity of optimal fair error: Ω(1/(γm)) to O(1/m) | Helped |
| Increase the dual-variable bound B (ablation) | DP-oracle-learner · Equalized Odds fair ERM | error bound | Hurt |
| Increase the dual-variable bound B (ablation) | DP-oracle-learner · Equalized Odds fair ERM | fairness violation bound | Helped |
| Require equalized false positive rates only, instead of equalized odds, and allow A-aware classifiers with group-indicator hypotheses (ablation) | DP-oracle-learner · Equalized false positive rate fair ERM | fairness violation bound | Helped |
Closest papers by meaning
- Differentially Private Post-Processing for Fair Regression (ICML 2024 · similarity 0.90)
- Oracle-Efficient Differentially Private Learning with Public Data (NeurIPS 2024 · similarity 0.88)
- Certification for Differentially Private Prediction in Gradient-Based Training (ICML 2025 · similarity 0.88)
- Instance-Specific Asymmetric Sensitivity in Differential Privacy (NeurIPS 2024 · similarity 0.87)
- Differentially Private Model Merging (NeurIPS 2026 · similarity 0.87)
- Can Private Machine Learning Be Fair? (AAAI 2025 · 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.