Strategic Conformal Prediction

Daniel Csillag, Claudio Jose Struchiner, Guilherme Tegoni Goedert
Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, PMLR 258:5122-5130, 2025.

Abstract

When a machine learning model is deployed, its predictions can alter its environment, as better informed agents strategize to suit their own interests. With such alterations in mind, existing approaches to uncertainty quantification break. In this work we propose a new framework, Strategic Conformal Prediction, which is capable of robust uncertainty quantification in such a setting. Strategic Conformal Prediction is backed by a series of theoretical guarantees spanning marginal coverage, training-conditional coverage, tightness and robustness to misspecification that hold in a distribution-free manner. Experimental analysis further validates our method, showing its remarkable effectiveness in face of arbitrary strategic alterations, whereas other methods break.

Cite this Paper

BibTeX
@InProceedings{pmlr-v258-csillag25a, title = {Strategic Conformal Prediction}, author = {Csillag, Daniel and Struchiner, Claudio Jose and Goedert, Guilherme Tegoni}, booktitle = {Proceedings of The 28th International Conference on Artificial Intelligence and Statistics}, pages = {5122--5130}, year = {2025}, editor = {Li, Yingzhen and Mandt, Stephan and Agrawal, Shipra and Khan, Emtiyaz}, volume = {258}, series = {Proceedings of Machine Learning Research}, month = {03--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v258/main/assets/csillag25a/csillag25a.pdf}, url = {https://proceedings.mlr.press/v258/csillag25a.html}, abstract = {When a machine learning model is deployed, its predictions can alter its environment, as better informed agents strategize to suit their own interests. With such alterations in mind, existing approaches to uncertainty quantification break. In this work we propose a new framework, Strategic Conformal Prediction, which is capable of robust uncertainty quantification in such a setting. Strategic Conformal Prediction is backed by a series of theoretical guarantees spanning marginal coverage, training-conditional coverage, tightness and robustness to misspecification that hold in a distribution-free manner. Experimental analysis further validates our method, showing its remarkable effectiveness in face of arbitrary strategic alterations, whereas other methods break.} }
Endnote
%0 Conference Paper %T Strategic Conformal Prediction %A Daniel Csillag %A Claudio Jose Struchiner %A Guilherme Tegoni Goedert %B Proceedings of The 28th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2025 %E Yingzhen Li %E Stephan Mandt %E Shipra Agrawal %E Emtiyaz Khan %F pmlr-v258-csillag25a %I PMLR %P 5122--5130 %U https://proceedings.mlr.press/v258/csillag25a.html %V 258 %X When a machine learning model is deployed, its predictions can alter its environment, as better informed agents strategize to suit their own interests. With such alterations in mind, existing approaches to uncertainty quantification break. In this work we propose a new framework, Strategic Conformal Prediction, which is capable of robust uncertainty quantification in such a setting. Strategic Conformal Prediction is backed by a series of theoretical guarantees spanning marginal coverage, training-conditional coverage, tightness and robustness to misspecification that hold in a distribution-free manner. Experimental analysis further validates our method, showing its remarkable effectiveness in face of arbitrary strategic alterations, whereas other methods break.
APA
Csillag, D., Struchiner, C.J. & Goedert, G.T.. (2025). Strategic Conformal Prediction. Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 258:5122-5130 Available from https://proceedings.mlr.press/v258/csillag25a.html.

Related Material