@InProceedings{marquesf2025,
title = {Stacked conformal prediction},
author = {Paulo C. {Marques F.}},
booktitle = {Proceedings of the Fourteenth Symposium on Conformal and Probabilistic Prediction with Applications},
pages = {305--316},
year = {2025},
editor = {Nguyen, Khuong An and Luo, Zhiyuan and Papadopoulos, Harris and L{\"o}fstr{\"o}m, Tuve and Carlsson, Lars and Bostr{\"o}m, Henrik},
volume = {266},
series = {Proceedings of Machine Learning Research},
month = {September},
publisher = {PMLR},
url = {https://proceedings.mlr.press/v266/marques25a.html}
}We consider the conformalization of a stacked ensemble of predictive models, showing that the potentially simple form of the meta-learner at the top of the stack enables a procedure with manageable computational cost that achieves approximate marginal validity without requiring the use of a separate calibration sample. Empirical results indicate that the method compares favorably to a standard inductive alternative.