Generalized Venn and Venn-Abers Calibration with Applications in Conformal Prediction

Lars van der Laan, Ahmed Alaa · 2025

Read Generalized Venn and Venn-Abers Calibration with Applications in Conformal Prediction by Lars van der Laan, Ahmed Alaa (2025) 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-2502-05676.txt.

Description

Lars van der Laan and Ahmed Alaa generalize a calibration method for prediction models in this 2025 paper. A model is calibrated when its stated confidence matches reality, and the classic distribution free tools only promise that match in the long run.

Their unified framework extends Venn and Venn-Abers calibration beyond yes or no classification to any problem set by a loss function. It turns a predictor that is calibrated on its own training data into one that, on finite samples, offers at least one honest point prediction inside a shrinking set. A multicalibration version aims the same guarantee at subpopulations, and for quantile loss it recovers conformal prediction as a special case.

Catalog record

CREATORLars van der Laan, Ahmed Alaa
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE62 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-05676

Browse the full library · About the Tower of Babel