Multi-scale Minimal Sufficient Representation Learning for Domain Generalization in Sleep Staging
Keywords: Domain generalization, Multi-scale, Contrastive learning, Information theory, Sleep staging
TL;DR: We propose a novel Multi-Scale Minimal Sufficient representation learning (MSMS) for sleep staging task that reduces domain-relevant information while preserving temporal and spectral information.
Abstract: Deep learning-based automatic sleep staging demonstrates strong performance as a promising solution for diagnosing sleep disorders. However, deep learning models often struggle to generalize on unseen subjects due to variability in physiological signals, resulting in degraded performance in out-of-distribution scenarios. To address this issue, domain generalization approaches have recently been studied actively to ensure generalized performance on unseen domains during the training. Among those techniques, contrastive learning has proven its validity in learning domain-invariant features by aligning samples of the same class across different domains. Despite its potential, many existing methods are insufficient for extracting truly domain-invariant representations, as they do not explicitly reduce domain-relevant information embedded in the features. In this paper, we argue that addressing superfluous information is a key to bridging the domain gap. Furthermore, existing methods often neglect the multi-scale nature of sleep signals, potentially missing important temporal and spectral characteristics. To address these limitations, we propose a novel Multi-Scale Minimal Sufficient representation learning (MSMS) framework, which effectively reduces domain-relevant information while preserving essential temporal and spectral features for sleep stage classification. We evaluate our method on publicly available sleep staging benchmark datasets, SleepEDF-20 and MASS. Experimental results demonstrate that our approach consistently outperforms state-of-the-art methods.
Supplementary Material: zip
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.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.
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: 9857
Loading