Saved in:
Bibliographic Details
Main Authors: Siahkali, Farbod, Verma, Ashwin, Gupta, Vijay
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.14913
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914332996009984
author Siahkali, Farbod
Verma, Ashwin
Gupta, Vijay
author_facet Siahkali, Farbod
Verma, Ashwin
Gupta, Vijay
contents Conformal prediction (CP) offers distribution-free marginal coverage guarantees under an exchangeability assumption, but these guarantees can fail if the data distribution shifts. We analyze the use of pseudo-calibration as a tool to counter this performance loss under a bounded label-conditional covariate shift model. Using tools from domain adaptation, we derive a lower bound on target coverage in terms of the source-domain loss of the classifier and a Wasserstein measure of the shift. Using this result, we provide a method to design pseudo-calibrated sets that inflate the conformal threshold by a slack parameter to keep target coverage above a prescribed level. Finally, we propose a source-tuned pseudo-calibration algorithm that interpolates between hard pseudo-labels and randomized labels as a function of classifier uncertainty. Numerical experiments show that our bounds qualitatively track pseudo-calibration behavior and that the source-tuned scheme mitigates coverage degradation under distribution shift while maintaining nontrivial prediction set sizes.
format Preprint
id arxiv_https___arxiv_org_abs_2602_14913
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Coverage Guarantees for Pseudo-Calibrated Conformal Prediction under Distribution Shift
Siahkali, Farbod
Verma, Ashwin
Gupta, Vijay
Machine Learning
Image and Video Processing
Conformal prediction (CP) offers distribution-free marginal coverage guarantees under an exchangeability assumption, but these guarantees can fail if the data distribution shifts. We analyze the use of pseudo-calibration as a tool to counter this performance loss under a bounded label-conditional covariate shift model. Using tools from domain adaptation, we derive a lower bound on target coverage in terms of the source-domain loss of the classifier and a Wasserstein measure of the shift. Using this result, we provide a method to design pseudo-calibrated sets that inflate the conformal threshold by a slack parameter to keep target coverage above a prescribed level. Finally, we propose a source-tuned pseudo-calibration algorithm that interpolates between hard pseudo-labels and randomized labels as a function of classifier uncertainty. Numerical experiments show that our bounds qualitatively track pseudo-calibration behavior and that the source-tuned scheme mitigates coverage degradation under distribution shift while maintaining nontrivial prediction set sizes.
title Coverage Guarantees for Pseudo-Calibrated Conformal Prediction under Distribution Shift
topic Machine Learning
Image and Video Processing
url https://arxiv.org/abs/2602.14913