OxEnsemble: Fair Ensembles for Low-Data Classification

Jonathan Rystrøm, Zihao Fu, Chris Russell
Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, PMLR 315:280-307, 2026.

Abstract

We address the problem of fair classification in settings where data is scarce and unbalanced across demographic groups. Such low-data regimes are common in domains like medical imaging, where false negatives can have fatal consequences. We propose a novel approach OxEnsemble for efficiently training ensembles and enforcing fairness in these low-data regimes. Unlike other approaches, we aggregate predictions across ensemble members, each trained to satisfy fairness constraints. By construction, OxEnsemble is both data-efficient – carefully reusing held-out data to enforce fairness reliably – and compute-efficient, requiring little more compute than used to fine-tune or evaluate an existing model. We validate this approach with new theoretical guarantees. Experimentally, our approach yields more consistent outcomes and stronger fairness-accuracy trade-offs than existing methods across multiple challenging medical imaging classification datasets.

Cite this Paper

BibTeX
@InProceedings{pmlr-v315-rystrom26a, title = {OxEnsemble: Fair Ensembles for Low-Data Classification}, author = {Rystr{\o}m, Jonathan and Fu, Zihao and Russell, Chris}, booktitle = {Proceedings of The 9th International Conference on Medical Imaging with Deep Learning}, pages = {280--307}, year = {2026}, editor = {Huo, Yuankai and Gao, Mingchen and Kuo, Chang-Fu and Jin, Yueming and Deng, Ruining}, volume = {315}, series = {Proceedings of Machine Learning Research}, month = {08--10 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v315/main/assets/rystrom26a/rystrom26a.pdf}, url = {https://proceedings.mlr.press/v315/rystrom26a.html}, abstract = {We address the problem of fair classification in settings where data is scarce and unbalanced across demographic groups. Such low-data regimes are common in domains like medical imaging, where false negatives can have fatal consequences. We propose a novel approach OxEnsemble for efficiently training ensembles and enforcing fairness in these low-data regimes. Unlike other approaches, we aggregate predictions across ensemble members, each trained to satisfy fairness constraints. By construction, OxEnsemble is both data-efficient – carefully reusing held-out data to enforce fairness reliably – and compute-efficient, requiring little more compute than used to fine-tune or evaluate an existing model. We validate this approach with new theoretical guarantees. Experimentally, our approach yields more consistent outcomes and stronger fairness-accuracy trade-offs than existing methods across multiple challenging medical imaging classification datasets.} }
Endnote
%0 Conference Paper %T OxEnsemble: Fair Ensembles for Low-Data Classification %A Jonathan Rystrøm %A Zihao Fu %A Chris Russell %B Proceedings of The 9th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2026 %E Yuankai Huo %E Mingchen Gao %E Chang-Fu Kuo %E Yueming Jin %E Ruining Deng %F pmlr-v315-rystrom26a %I PMLR %P 280--307 %U https://proceedings.mlr.press/v315/rystrom26a.html %V 315 %X We address the problem of fair classification in settings where data is scarce and unbalanced across demographic groups. Such low-data regimes are common in domains like medical imaging, where false negatives can have fatal consequences. We propose a novel approach OxEnsemble for efficiently training ensembles and enforcing fairness in these low-data regimes. Unlike other approaches, we aggregate predictions across ensemble members, each trained to satisfy fairness constraints. By construction, OxEnsemble is both data-efficient – carefully reusing held-out data to enforce fairness reliably – and compute-efficient, requiring little more compute than used to fine-tune or evaluate an existing model. We validate this approach with new theoretical guarantees. Experimentally, our approach yields more consistent outcomes and stronger fairness-accuracy trade-offs than existing methods across multiple challenging medical imaging classification datasets.
APA
Rystrøm, J., Fu, Z. & Russell, C.. (2026). OxEnsemble: Fair Ensembles for Low-Data Classification. Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 315:280-307 Available from https://proceedings.mlr.press/v315/rystrom26a.html.

Related Material