Rethinking pseudo-labeling: Data-centric insights improve semi-supervised learning

Nabeel Seedat, Nicolas Huynh, Fergus Imrie, Mihaela van der Schaar

Primary Area: unsupervised, self-supervised, semi-supervised, and supervised representation learning

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Keywords: pseudo-labeling, semi-supervised learning, data-centric AI

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.

TL;DR: Labeled data is unrealistically assumed to be gold-standard in semi-supervised learning. We show data-centric insights can improve pseudo-labeling in real-world settings.

Abstract: Pseudo-labeling is a popular semi-supervised learning technique to leverage unlabeled data when labeled samples are scarce. The generation and selection of pseudo-labels heavily rely on labeled data. Existing approaches implicitly assume that the labeled data is gold standard and “perfect”. However, this can be violated in reality with issues such as mislabeling or ambiguity. We address this overlooked aspect and show the importance of investigating labeled data quality to improve any pseudo-labeling method. Specifically, we introduce a novel data characterization and selection framework called DIPS to extend pseudo-labeling. We select useful labeled and pseudo-labeled samples via analysis of learning dynamics. We empirically demonstrate that DIPS improves the performance of various pseudo-labeling methods on real-world datasets across multiple modalities, including tabular and images, with minimal computational overhead. Additionally, DIPS improves data efficiency and reduces the performance distinctions between different pseudo-labelers. Overall, we highlight the significant benefits of a data-centric rethinking of pseudo-labeling in real-world settings.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 3725