Research map / Learning theory and algorithms / Fairness
Multicalibration: Calibration for the (Computationally-Identifiable) Masses
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 |
|---|---|---|---|
| Enforce multicalibration (calibration on every subpopulation in a class C) on a predictor after learning | any base predictor h · theoretical (squared error over X) | squared error relative to best-in-class predictor: < 6α worse than h* | No clear effect |
| Post-process a predictor to be multicalibrated on the categories of h (and class C) | predictor h · theoretical (squared error) | squared prediction error: at most 6α worse than h on each category | No clear effect |
| Replace balance/statistical-parity-based group fairness with calibration-based multicalibration | fair predictor design · conceptual comparison | fairness-accuracy tradeoff | Helped |
| Use guess-and-check oracle based on differential privacy to answer adaptive statistical queries | multicalibration learning algorithm · finite sample generalization | sample complexity | Helped |
Closest papers by meaning
- Fair Risk Control: A Generalized Framework for Calibrating Multi-group Fairness Risks (ICML 2024 · similarity 0.89)
- Who’s the (Multi-)Fairest of Them All: Rethinking Interpolation-Based Data Augmentation Through the Lens of Multicalibration (AAAI 2025 · similarity 0.88)
- Multiaccuracy and Multicalibration via Proxy Groups (ICML 2025 · similarity 0.88)
- On the Burden of Achieving Fairness in Conformal Prediction (NeurIPS 2026 · similarity 0.87)
- Improved and Oracle-Efficient Online $\ell_1$-Multicalibration (ICML 2025 · similarity 0.87)
- Bridging Multicalibration and Out-of-distribution Generalization Beyond Covariate Shift (NeurIPS 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.