Conformal Risk Control under Non-Monotone Losses: Theory and Finite-Sample Guarantees
Tareq Aldirawi, Yun Li, Wenge Guo · 2026
Read Conformal Risk Control under Non-Monotone Losses: Theory and Finite-Sample Guarantees by Tareq Aldirawi, Yun Li, Wenge Guo (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-01502.txt.
Description
Aldirawi and colleagues study conformal risk control under non-monotone loss functions when the parameter is chosen from a finite grid. They show validity without monotonicity depends on calibration sample size versus grid resolution. They establish a finite guarantee for losses: excess risk above the target scales on the order of the square root of log grid size over n, where n is the calibration. A lower bound shows this rate is minimax optimal. Methods accounting for finite uncertainty achieve more stable risk control than approaches based on monotonicity transformations.
Catalog record
| CREATOR | Tareq Aldirawi, Yun Li, Wenge Guo |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 81599 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.01502