Retrieving predictive densities from conformal predictive distributions
Abstract
Conformal prediction intervals are used to construct marginally well-calibrated regions around point predictions. Recent work has focused on extending these ideas to conformal predictive distributions, whose marginal probability integral transform (PIT) follows a Unif[0, 1] distribution. In this work, we investigate how to recover predictive densities from conformal predictive distributions. We extend the theory of conformal predictive distributions by proving asymptotic marginal validity for a tail-corrected version of conformal predictive distributions. We propose a method called quantile matching, which preserves an upper bound on the deviation of the marginal PIT from uniformity. Furthermore, we show that the distribution induced by quantile matching is equivalent to the crisp version of conformal predictive distributions when the number of quantiles equals the size of the calibration set. For the recovery of conformal densities, we construct a fidelity-constrained bandwidth optimization for kernel smoothing that preserves asymptotic marginal validity and has a closed-form solution. We apply and compare our proposed methodology on a large simulated real estate transactions dataset based on the Hierarchical Trend Model.
Related
- Conformal Predictive Distributions for Contextual Order Fulfillment Optimization
- Conformal ordering: Distribution-free predictive inference for orders and rank
- Calibrated Predictive Distributions from Sample-Based Generators
- Coherence is not enough: Aggregation constraints across predictive distributions, forecasts and decisions
- SCOPE Before You Sample: Conformal Scores Guide Posterior Exploration for Fast and Calibrated Bayesian Predictive Inference
BibTeX
@misc{tudor2026retrieving,
title = {{Retrieving predictive densities from conformal predictive distributions}},
author = {Dan Andrei Tudor and Dorota Kurowicka and Balint Negyesi and Valerii Zoller},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.23350},
eprint = {2609.23350},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.23350},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Retrieving predictive densities from conformal predictive distributions
AU - Dan Andrei Tudor
AU - Dorota Kurowicka
AU - Balint Negyesi
AU - Valerii Zoller
PY - 2026
DA - 2026/09/20
PB - arXiv
DO - 10.48550/arxiv.2609.23350
UR - https://arxiv.org/abs/2609.23350
AB - Conformal prediction intervals are used to construct marginally well-calibrated regions around point predictions. Recent work has focused on extending these ideas to conformal predictive distributions, whose marginal probability integral transform (PIT) follows a Unif[0, 1] distribution. In this work, we investigate how to recover predictive densities from conformal predictive distributions. We extend the theory of conformal predictive distributions by proving asymptotic marginal validity for a tail-corrected version of conformal predictive distributions. We propose a method called quantile matching, which preserves an upper bound on the deviation of the marginal PIT from uniformity. Furthermore, we show that the distribution induced by quantile matching is equivalent to the crisp version of conformal predictive distributions when the number of quantiles equals the size of the calibration set. For the recovery of conformal densities, we construct a fidelity-constrained bandwidth optimization for kernel smoothing that preserves asymptotic marginal validity and has a closed-form solution. We apply and compare our proposed methodology on a large simulated real estate transactions dataset based on the Hierarchical Trend Model.
ER -