Research map / Learning theory and algorithms / Fairness
Leveraging Labeled and Unlabeled Data for Consistent Fair Binary Classification
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 |
|---|---|---|---|
| Apply group-dependent threshold calibration to RF probabilities | RF · Arrhythmia | DEO: $0.09 \pm 0.02$ -> $0.03 \pm 0.01$ | Helped |
| Apply group-dependent threshold calibration to RF probabilities | RF · COMPAS | DEO: $0.11 \pm 0.02$ -> $0.04 \pm 0.02$ | Helped |
| Apply group-dependent threshold calibration to RF probabilities | RF · Adult | DEO: $0.12$ -> $0.03$ | Helped |
| Apply group-dependent threshold calibration to RF probabilities | RF · German | DEO: $0.09 \pm 0.04$ -> $0.02 \pm 0.02$ | Helped |
| Apply group-dependent threshold calibration to RF probabilities | RF · Drug | DEO: $0.29 \pm 0.02$ -> $0.04 \pm 0.02$ | Helped |
| Apply group-dependent threshold calibration to RF probabilities | RF · Arrhythmia | ACC: $0.83 \pm 0.03$ -> $0.81 \pm 0.04$ | Hurt |
| Apply group-dependent threshold calibration to RF probabilities | RF · COMPAS | ACC: $0.77 \pm 0.02$ -> $0.76 \pm 0.02$ | Hurt |
| Apply group-dependent threshold calibration to RF probabilities | RF · Adult | ACC: $0.86$ -> $0.85$ | Hurt |
| Apply group-dependent threshold calibration to RF probabilities | RF · German | ACC: $0.78 \pm 0.02$ -> $0.77 \pm 0.02$ | Hurt |
| Apply group-dependent threshold calibration to RF probabilities | RF · Drug | ACC: $0.86 \pm 0.01$ -> $0.83 \pm 0.01$ | Hurt |
| Increase unlabeled sample from 1/10 to 8/10 while keeping labeled data at 1/10 (ablation) | RF+Ours · COMPAS | ACC: $0.68 \pm 0.03$ -> $0.71 \pm 0.02$ | Helped |
| Increase unlabeled sample from 1/10 to 8/10 while keeping labeled data at 1/10 (ablation) | RF+Ours · COMPAS | DEO: $0.07 \pm 0.02$ -> $0.05 \pm 0.01$ | Helped |
Closest papers by meaning
- A Reductions Approach to Fair Classification (ICML 2018 · similarity 0.92)
- Bayes-Optimal Fair Classification with Multiple Sensitive Features (AAAI 2026 · similarity 0.92)
- Post-hoc bias scoring is optimal for fair classification (ICLR 2024 · similarity 0.90)
- FROC: Building Fair ROC from a Trained Classifier (AAAI 2025 · similarity 0.88)
- Efficient Fairness-Performance Pareto Front Computation (NeurIPS 2025 · similarity 0.87)
- Arbitrariness and Social Prediction: The Confounding Role of Variance in Fair Classification (AAAI 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.