Optimal Learning from Label Proportions with General Loss Functions
Lorne Applebaum, Travis Dick, Claudio Gentile, Haim Kaplan, Tomer Koren
Abstract
Motivated by problems in online advertising, we address the task of Learning from Label Proportions (LLP). We introduce a novel and versatile low-variance debiasing methodology to learn from aggregate label information, significantly advancing the state of the art in LLP. Our debiasing approach exhibits remarkable flexibility, seamlessly accommodating a broad spectrum of practically relevant loss functions across both binary and multi-class classification settings. By carefully combining our estimators with standard techniques, we improve sample complexity guarantees for a large class of losses of practical relevance. We also empirically validate the efficacy of our proposed approach across a diverse array of benchmark datasets, demonstrating compelling empirical advantages over standard baselines.
BibTeX
@inproceedings{
applebaum2026optimal,
title={Optimal Learning from Label Proportions with General Loss Functions},
author={Lorne Applebaum and Travis Dick and Claudio Gentile and Haim Kaplan and Tomer Koren},
booktitle={Forty-third International Conference on Machine Learning},
year={2026},
url={https://openreview.net/forum?id=GLNiB1ePxm}
}