Conformal Prediction with Stable Measurement Mechanisms
Abstract
In applications such as air-quality monitoring, inexpensive sensor readings can help predict pollutant concentrations measured by a separate reference instrument. We study how to adapt conformal prediction intervals to a new population when reference outcomes are available in the source data, but only sensor readings and context, such as weather conditions, are available in the target. We assume that the measurement distribution at a given outcome and context remains stable, while context frequencies and outcome distributions may change. This model is a particular form of causal mechanism shift. Its statistical restriction has been studied as sparse joint shift. Using established marginal matching, predictive adjustment, and weighted conformal calibration, we examine what this restriction contributes to coverage and interval length for continuous outcomes. Gaussian analyses show how measurement stability can identify outcome changes and how retaining context can shorten prediction intervals. For Gaussian models with known source mechanisms, we derive exact coverage formulas in the large-calibration limit while keeping the unlabeled target sample finite. These formulas explain when sharing a correction across contexts is preferable to fitting separate corrections, and why estimating outcome variability is particularly sensitive to weakly informative measurements. Experiments on synthetic Gaussian data show a coverage gain over covariate weighting at the same mean length and a length reduction of about 8% relative to pooled label adjustment at comparable coverage. On the Air Quality dataset, controlled shifts yield coverage gains with longer intervals, while spring-to-autumn and year-over-year transfers show substantial undercoverage.
Related
- Beyond Imputation: Mask-Adaptive Conformal Prediction via Tree Embeddings on General Missing Data Mechanisms
- Conformal structures and rigidity of complete stable minimal hypersurfaces
- Generalized Hyperrectangular Prediction Sets in Multivariate Conformal Prediction
- Conformal Prediction under Partial Verification
- Multi-Variable Conformal Prediction: Optimizing Prediction Sets using Multiple Calibration Variables
BibTeX
@misc{choi2026conformal,
title = {{Conformal Prediction with Stable Measurement Mechanisms}},
author = {Seungjin Choi},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.34204},
eprint = {2610.34204},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.34204},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Conformal Prediction with Stable Measurement Mechanisms
AU - Seungjin Choi
PY - 2026
DA - 2026/10/25
PB - arXiv
DO - 10.48550/arxiv.2610.34204
UR - https://arxiv.org/abs/2610.34204
AB - In applications such as air-quality monitoring, inexpensive sensor readings can help predict pollutant concentrations measured by a separate reference instrument. We study how to adapt conformal prediction intervals to a new population when reference outcomes are available in the source data, but only sensor readings and context, such as weather conditions, are available in the target. We assume that the measurement distribution at a given outcome and context remains stable, while context frequencies and outcome distributions may change. This model is a particular form of causal mechanism shift. Its statistical restriction has been studied as sparse joint shift. Using established marginal matching, predictive adjustment, and weighted conformal calibration, we examine what this restriction contributes to coverage and interval length for continuous outcomes. Gaussian analyses show how measurement stability can identify outcome changes and how retaining context can shorten prediction intervals. For Gaussian models with known source mechanisms, we derive exact coverage formulas in the large-calibration limit while keeping the unlabeled target sample finite. These formulas explain when sharing a correction across contexts is preferable to fitting separate corrections, and why estimating outcome variability is particularly sensitive to weakly informative measurements. Experiments on synthetic Gaussian data show a coverage gain over covariate weighting at the same mean length and a length reduction of about 8% relative to pooled label adjustment at comparable coverage. On the Air Quality dataset, controlled shifts yield coverage gains with longer intervals, while spring-to-autumn and year-over-year transfers show substantial undercoverage.
ER -