CoCoNet: Convex Conformal Network for Multi-Dimensional Uncertainty Regions
Abstract
Conformal prediction (CP) provides distribution-free coverage guarantees under exchangeability, but the efficiency of its prediction regions depends critically on the choice of the conformity score. This is particularly important in multivariate regression, where dependencies across output dimensions induce complex joint uncertainty structures. This paper introduces CoCoNet (Convex Conformal Network) for learning convex conformity scores that induce flexible, data-adaptive prediction regions. CoCoNet learns the score from held-out data by optimizing prediction-region efficiency while enforcing convexity with respect to the residual. CoCoNet then applies split-conformal calibration to retain finite-sample coverage guarantees. The paper presents two variants of the method: CoCoNetSpline, based on piecewise-linear scores, and CoCoNetICNN, based on Input Convex Neural Networks. Experiments compare CoCoNet with other CP baselines over synthetic and real-world datasets with output dimensions ranging from 2 to 16. The results demonstrate the flexibility of CoCoNet in learning different response types. Across 8 models and 30 multi-dimensional regression tasks, CoCoNetICNN and CoCoNetSpline−ℓ1 achieve the best efficiencies with average ranks of 1.45 and 2.75, respectively. In particular, CoCoNetICNN ranks first on 22 tasks out of 30 in terms of efficiency.
Related
- Conformal prediction regions for dependent censored lifetimes
- SPACR: Single-Pass Adaptive Training of Uncertainty-Aware Conformal Regressors
- Common-Center Geometry and Certified Radial Reconstruction for Energy-Form Full Conformal Regions
- Conformal Uncertainty Quantification Guarantees for Neural Operators
- CASCADE Conformal Prediction: Uncertainty-Adaptive Prediction Intervals for Two-Stage Clinical Decision Support
BibTeX
@misc{moradi2026coconet,
title = {{CoCoNet: Convex Conformal Network for Multi-Dimensional Uncertainty Regions}},
author = {Alireza Moradi and Thomas Bruys and Pascal Van Hentenryck and Reza Zandehshahvar},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.24246},
eprint = {2610.24246},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.24246},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - CoCoNet: Convex Conformal Network for Multi-Dimensional Uncertainty Regions
AU - Alireza Moradi
AU - Thomas Bruys
AU - Pascal Van Hentenryck
AU - Reza Zandehshahvar
PY - 2026
DA - 2026/10/28
PB - arXiv
DO - 10.48550/arxiv.2610.24246
UR - https://arxiv.org/abs/2610.24246
AB - Conformal prediction (CP) provides distribution-free coverage guarantees under exchangeability, but the efficiency of its prediction regions depends critically on the choice of the conformity score. This is particularly important in multivariate regression, where dependencies across output dimensions induce complex joint uncertainty structures. This paper introduces CoCoNet (Convex Conformal Network) for learning convex conformity scores that induce flexible, data-adaptive prediction regions. CoCoNet learns the score from held-out data by optimizing prediction-region efficiency while enforcing convexity with respect to the residual. CoCoNet then applies split-conformal calibration to retain finite-sample coverage guarantees. The paper presents two variants of the method: CoCoNet_Spline, based on piecewise-linear scores, and CoCoNet_ICNN, based on Input Convex Neural Networks. Experiments compare CoCoNet with other CP baselines over synthetic and real-world datasets with output dimensions ranging from 2 to 16. The results demonstrate the flexibility of CoCoNet in learning different response types. Across 8 models and 30 multi-dimensional regression tasks, CoCoNet_ICNN and CoCoNet_Spline−ℓ_1 achieve the best efficiencies with average ranks of 1.45 and 2.75, respectively. In particular, CoCoNet_ICNN ranks first on 22 tasks out of 30 in terms of efficiency.
ER -