A Universal Unbiased Method for Classification from Aggregate Observations

Zixi Wei, Lei Feng, Bo Han, Tongliang Liu, Gang Niu, Xiaofeng Zhu, Heng Tao Shen
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:36804-36820, 2023.

Abstract

In conventional supervised classification, true labels are required for individual instances. However, it could be prohibitive to collect the true labels for individual instances, due to privacy concerns or unaffordable annotation costs. This motivates the study on classification from aggregate observations (CFAO), where the supervision is provided to groups of instances, instead of individual instances. CFAO is a generalized learning framework that contains various learning problems, such as multiple-instance learning and learning from label proportions. The goal of this paper is to present a novel universal method of CFAO, which holds an unbiased estimator of the classification risk for arbitrary losses—previous research failed to achieve this goal. Practically, our method works by weighing the importance of each instance and each label in the group, which provides purified supervision for the classifier to learn. Theoretically, our proposed method not only guarantees the risk consistency due to the unbiased risk estimator but also can be compatible with arbitrary losses. Extensive experiments on various problems of CFAO demonstrate the superiority of our proposed method.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-wei23a, title = {A Universal Unbiased Method for Classification from Aggregate Observations}, author = {Wei, Zixi and Feng, Lei and Han, Bo and Liu, Tongliang and Niu, Gang and Zhu, Xiaofeng and Shen, Heng Tao}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {36804--36820}, 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/wei23a/wei23a.pdf}, url = {https://proceedings.mlr.press/v202/wei23a.html}, abstract = {In conventional supervised classification, true labels are required for individual instances. However, it could be prohibitive to collect the true labels for individual instances, due to privacy concerns or unaffordable annotation costs. This motivates the study on classification from aggregate observations (CFAO), where the supervision is provided to groups of instances, instead of individual instances. CFAO is a generalized learning framework that contains various learning problems, such as multiple-instance learning and learning from label proportions. The goal of this paper is to present a novel universal method of CFAO, which holds an unbiased estimator of the classification risk for arbitrary losses—previous research failed to achieve this goal. Practically, our method works by weighing the importance of each instance and each label in the group, which provides purified supervision for the classifier to learn. Theoretically, our proposed method not only guarantees the risk consistency due to the unbiased risk estimator but also can be compatible with arbitrary losses. Extensive experiments on various problems of CFAO demonstrate the superiority of our proposed method.} }
Endnote
%0 Conference Paper %T A Universal Unbiased Method for Classification from Aggregate Observations %A Zixi Wei %A Lei Feng %A Bo Han %A Tongliang Liu %A Gang Niu %A Xiaofeng Zhu %A Heng Tao Shen %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-wei23a %I PMLR %P 36804--36820 %U https://proceedings.mlr.press/v202/wei23a.html %V 202 %X In conventional supervised classification, true labels are required for individual instances. However, it could be prohibitive to collect the true labels for individual instances, due to privacy concerns or unaffordable annotation costs. This motivates the study on classification from aggregate observations (CFAO), where the supervision is provided to groups of instances, instead of individual instances. CFAO is a generalized learning framework that contains various learning problems, such as multiple-instance learning and learning from label proportions. The goal of this paper is to present a novel universal method of CFAO, which holds an unbiased estimator of the classification risk for arbitrary losses—previous research failed to achieve this goal. Practically, our method works by weighing the importance of each instance and each label in the group, which provides purified supervision for the classifier to learn. Theoretically, our proposed method not only guarantees the risk consistency due to the unbiased risk estimator but also can be compatible with arbitrary losses. Extensive experiments on various problems of CFAO demonstrate the superiority of our proposed method.
APA
Wei, Z., Feng, L., Han, B., Liu, T., Niu, G., Zhu, X. & Shen, H.T.. (2023). A Universal Unbiased Method for Classification from Aggregate Observations. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:36804-36820 Available from https://proceedings.mlr.press/v202/wei23a.html.

Related Material