Fast Online Value-Maximizing Prediction Sets with Conformal Cost Control

Zhen Lin, Shubhendu Trivedi, Cao Xiao, Jimeng Sun
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:21182-21203, 2023.

Abstract

Many real-world multi-label prediction problems involve set-valued predictions that must satisfy specific requirements dictated by downstream usage. We focus on a typical scenario where such requirements, separately encoding value and cost, compete with each other. For instance, a hospital might expect a smart diagnosis system to capture as many severe, often co-morbid, diseases as possible (the value), while maintaining strict control over incorrect predictions (the cost). We present a general pipeline, dubbed as FavMac, to maximize the value while controlling the cost in such scenarios. FavMac can be combined with almost any multi-label classifier, affording distribution-free theoretical guarantees on cost control. Moreover, unlike prior works, FavMac can handle real-world large-scale applications via a carefully designed online update mechanism, which is of independent interest. Our methodological and theoretical contributions are supported by experiments on several healthcare tasks and synthetic datasets - FavMac furnishes higher value compared with several variants and baselines while maintaining strict cost control.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-lin23j, title = {Fast Online Value-Maximizing Prediction Sets with Conformal Cost Control}, author = {Lin, Zhen and Trivedi, Shubhendu and Xiao, Cao and Sun, Jimeng}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {21182--21203}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/lin23j/lin23j.pdf}, url = {https://proceedings.mlr.press/v202/lin23j.html}, abstract = {Many real-world multi-label prediction problems involve set-valued predictions that must satisfy specific requirements dictated by downstream usage. We focus on a typical scenario where such requirements, separately encoding value and cost, compete with each other. For instance, a hospital might expect a smart diagnosis system to capture as many severe, often co-morbid, diseases as possible (the value), while maintaining strict control over incorrect predictions (the cost). We present a general pipeline, dubbed as FavMac, to maximize the value while controlling the cost in such scenarios. FavMac can be combined with almost any multi-label classifier, affording distribution-free theoretical guarantees on cost control. Moreover, unlike prior works, FavMac can handle real-world large-scale applications via a carefully designed online update mechanism, which is of independent interest. Our methodological and theoretical contributions are supported by experiments on several healthcare tasks and synthetic datasets - FavMac furnishes higher value compared with several variants and baselines while maintaining strict cost control.} }
Endnote
%0 Conference Paper %T Fast Online Value-Maximizing Prediction Sets with Conformal Cost Control %A Zhen Lin %A Shubhendu Trivedi %A Cao Xiao %A Jimeng Sun %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-lin23j %I PMLR %P 21182--21203 %U https://proceedings.mlr.press/v202/lin23j.html %V 202 %X Many real-world multi-label prediction problems involve set-valued predictions that must satisfy specific requirements dictated by downstream usage. We focus on a typical scenario where such requirements, separately encoding value and cost, compete with each other. For instance, a hospital might expect a smart diagnosis system to capture as many severe, often co-morbid, diseases as possible (the value), while maintaining strict control over incorrect predictions (the cost). We present a general pipeline, dubbed as FavMac, to maximize the value while controlling the cost in such scenarios. FavMac can be combined with almost any multi-label classifier, affording distribution-free theoretical guarantees on cost control. Moreover, unlike prior works, FavMac can handle real-world large-scale applications via a carefully designed online update mechanism, which is of independent interest. Our methodological and theoretical contributions are supported by experiments on several healthcare tasks and synthetic datasets - FavMac furnishes higher value compared with several variants and baselines while maintaining strict cost control.
APA
Lin, Z., Trivedi, S., Xiao, C. & Sun, J.. (2023). Fast Online Value-Maximizing Prediction Sets with Conformal Cost Control. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:21182-21203 Available from https://proceedings.mlr.press/v202/lin23j.html.

Related Material