Research map / Learning theory and algorithms / Sample complexity
Sample Complexity of Uniform Convergence for Multicalibration
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 the finite-class uniform convergence analysis | Multicalibration error for finite predictor classes · Finite predictor classes | sample complexity dependence on generalization error: \epsilon^{-3} -> \epsilon^{-2} | Helped |
| Use the finite-class uniform convergence analysis | Multicalibration error for finite predictor classes · Finite predictor classes | sample complexity dependence on group mass: \gamma^{-3/2} -> \gamma^{-1} | Helped |
| Use the finite-class uniform convergence analysis | Multicalibration error for finite predictor classes · Finite predictor classes | sample complexity dependence on category frequency: \psi^{-3/2} -> \psi^{-1} | Helped |
| Use the graph-dimension uniform convergence analysis | Multicalibration error for infinite predictor classes · Infinite classes with finite graph dimension | sample complexity dependence on category frequency: \psi^{-3/2} -> \psi^{-2} | Mixed |
| Bound calibration error directly instead of through prediction-loss gap | Multicalibration generalization · Agnostic learning | calibration error guarantee | Helped |
| Extend the analysis beyond a differentially private algorithm and finite domains | Multicalibration uniform convergence · Continuous domains | applicability | 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)
- Improved Bounds for Swap Multicalibration and Swap Omniprediction (NeurIPS 2025 · similarity 0.88)
- Improved and Oracle-Efficient Online $\ell_1$-Multicalibration (ICML 2025 · similarity 0.87)
- Multiaccuracy and Multicalibration via Proxy Groups (ICML 2025 · similarity 0.87)
- On the Maximal Local Disparity of Fairness-Aware Classifiers (ICML 2024 · similarity 0.86)
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.