Semi-Verified PAC Learning from the Crowd

Shiwei Zeng, Jie Shen
Proceedings of The 26th International Conference on Artificial Intelligence and Statistics, PMLR 206:505-522, 2023.

Abstract

We study the problem of crowdsourced PAC learning of threshold functions. This is a challenging problem and only recently have query-efficient algorithms been established under the assumption that a noticeable fraction of the workers are perfect. In this work, we investigate a more challenging case where the majority may behave adversarially and the rest behave as the Massart noise – a significant generalization of the perfectness assumption. We show that under the semi-verified model of Charikar et al. (2017), where we have (limited) access to a trusted oracle who always returns correct annotations, it is possible to PAC learn the underlying hypothesis class with a manageable amount of label queries. Moreover, we show that the labeling cost can be drastically mitigated via the more easily obtained comparison queries. Orthogonal to recent developments in semi-verified or list-decodable learning that crucially rely on data distributional assumptions, our PAC guarantee holds by exploring the wisdom of the crowd.

Cite this Paper

BibTeX
@InProceedings{pmlr-v206-zeng23a, title = {Semi-Verified PAC Learning from the Crowd}, author = {Zeng, Shiwei and Shen, Jie}, booktitle = {Proceedings of The 26th International Conference on Artificial Intelligence and Statistics}, pages = {505--522}, year = {2023}, editor = {Ruiz, Francisco and Dy, Jennifer and van de Meent, Jan-Willem}, volume = {206}, series = {Proceedings of Machine Learning Research}, month = {25--27 Apr}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v206/zeng23a/zeng23a.pdf}, url = {https://proceedings.mlr.press/v206/zeng23a.html}, abstract = {We study the problem of crowdsourced PAC learning of threshold functions. This is a challenging problem and only recently have query-efficient algorithms been established under the assumption that a noticeable fraction of the workers are perfect. In this work, we investigate a more challenging case where the majority may behave adversarially and the rest behave as the Massart noise – a significant generalization of the perfectness assumption. We show that under the semi-verified model of Charikar et al. (2017), where we have (limited) access to a trusted oracle who always returns correct annotations, it is possible to PAC learn the underlying hypothesis class with a manageable amount of label queries. Moreover, we show that the labeling cost can be drastically mitigated via the more easily obtained comparison queries. Orthogonal to recent developments in semi-verified or list-decodable learning that crucially rely on data distributional assumptions, our PAC guarantee holds by exploring the wisdom of the crowd.} }
Endnote
%0 Conference Paper %T Semi-Verified PAC Learning from the Crowd %A Shiwei Zeng %A Jie Shen %B Proceedings of The 26th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2023 %E Francisco Ruiz %E Jennifer Dy %E Jan-Willem van de Meent %F pmlr-v206-zeng23a %I PMLR %P 505--522 %U https://proceedings.mlr.press/v206/zeng23a.html %V 206 %X We study the problem of crowdsourced PAC learning of threshold functions. This is a challenging problem and only recently have query-efficient algorithms been established under the assumption that a noticeable fraction of the workers are perfect. In this work, we investigate a more challenging case where the majority may behave adversarially and the rest behave as the Massart noise – a significant generalization of the perfectness assumption. We show that under the semi-verified model of Charikar et al. (2017), where we have (limited) access to a trusted oracle who always returns correct annotations, it is possible to PAC learn the underlying hypothesis class with a manageable amount of label queries. Moreover, we show that the labeling cost can be drastically mitigated via the more easily obtained comparison queries. Orthogonal to recent developments in semi-verified or list-decodable learning that crucially rely on data distributional assumptions, our PAC guarantee holds by exploring the wisdom of the crowd.
APA
Zeng, S. & Shen, J.. (2023). Semi-Verified PAC Learning from the Crowd. Proceedings of The 26th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 206:505-522 Available from https://proceedings.mlr.press/v206/zeng23a.html.

Related Material