Research map / Learning theory and algorithms / Fairness
On Fairness and Calibration
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 |
|---|---|---|---|
| Require calibration and equalized odds across groups with different base rates | Probabilistic classifiers · Groups with unequal base rates | false-positive and false-negative rates | Hurt |
| Add calibration while equalizing generalized false-negative rates | Income prediction (Adult Dataset) · Adult Dataset | generalized false-positive rate disparity | Hurt |
| Add calibration while equalizing weighted false-positive and false-negative cost | Health prediction (Heart Dataset) · Heart Dataset | equalized cost | No clear effect |
| Add calibration while equalizing generalized false-positive rates | COMPAS recidivism classifiers · Criminal recidivism prediction | feasibility of equalized false-positive rates | Hurt |
| Use training-time Equalized Odds constraints | COMPAS recidivism classifiers · Criminal recidivism prediction | false-positive and false-negative rate parity | Helped |
| Use post-processing to achieve Equalized Odds | COMPAS recidivism classifiers · Criminal recidivism prediction | false-positive and false-negative rate parity | Helped |
| Add a second distinct equal-cost constraint while retaining calibration | Calibrated classifiers for groups with unequal base rates · Groups with unequal base rates | false-positive and false-negative rates | Hurt |
Closest papers by meaning
- Multicalibration: Calibration for the (Computationally-Identifiable) Masses (ICML 2018 · similarity 0.91)
- Bounding the fairness and accuracy of classifiers from population statistics (ICML 2020 · similarity 0.90)
- Fairness via Independence: A General Regularization Framework for Machine Learning (ICLR 2026 · similarity 0.89)
- Sample Complexity of Uniform Convergence for Multicalibration (NeurIPS 2020 · similarity 0.89)
- Who’s the (Multi-)Fairest of Them All: Rethinking Interpolation-Based Data Augmentation Through the Lens of Multicalibration (AAAI 2025 · similarity 0.89)
- Accuracy vs. Accuracy: Computational Tradeoffs Between Classification Rates and Utility (NeurIPS 2026 · similarity 0.88)
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.