Self-Supervised Dynamical System Representations for Physiological Time-Series

Yenho Chen, Maxwell A Xu, James Matthew Rehg, Christopher John Rozell
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:17714-17733, 2026.

Abstract

Self-supervised learning for physiological time-series aims to captures the identity of the underlying dynamical process while filtering irrelevant noise. However, existing approaches may obscure the clinical semantics important for downstream transferability. Weakly constrained pretext tasks (i.e. contrastive learning, MAE) may incorrectly ignore the underlying dynamical structure, while structurally constrained models (i.e. SVAEs) are unable to selectively filter sample-specific noise. To bridge this gap, we propose ${\bf PULSE}$, a novel pretraining objective that simultaneously preserves dynamical relationships important to physiological time-series while selectively removing irrelevant noise. We achieve this by formulating a dynamical systems model to identify transferable and non-transferable information between time-series windows, and target the former through a novel cross-reconstruction objective. We establish theory that provides conditions for when transferrable information is recovered, and empirically validate it through synthetic experiments. On several real-world datasets, PULSE effectively distinguishes clinical semantic classes, increases label efficiency, and improves transfer learning performance.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-chen26fk, title = {Self-Supervised Dynamical System Representations for Physiological Time-Series}, author = {Chen, Yenho and Xu, Maxwell A and Rehg, James Matthew and Rozell, Christopher John}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {17714--17733}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/chen26fk/chen26fk.pdf}, url = {https://proceedings.mlr.press/v306/chen26fk.html}, abstract = {Self-supervised learning for physiological time-series aims to captures the identity of the underlying dynamical process while filtering irrelevant noise. However, existing approaches may obscure the clinical semantics important for downstream transferability. Weakly constrained pretext tasks (i.e. contrastive learning, MAE) may incorrectly ignore the underlying dynamical structure, while structurally constrained models (i.e. SVAEs) are unable to selectively filter sample-specific noise. To bridge this gap, we propose ${\bf PULSE}$, a novel pretraining objective that simultaneously preserves dynamical relationships important to physiological time-series while selectively removing irrelevant noise. We achieve this by formulating a dynamical systems model to identify transferable and non-transferable information between time-series windows, and target the former through a novel cross-reconstruction objective. We establish theory that provides conditions for when transferrable information is recovered, and empirically validate it through synthetic experiments. On several real-world datasets, PULSE effectively distinguishes clinical semantic classes, increases label efficiency, and improves transfer learning performance.} }
Endnote
%0 Conference Paper %T Self-Supervised Dynamical System Representations for Physiological Time-Series %A Yenho Chen %A Maxwell A Xu %A James Matthew Rehg %A Christopher John Rozell %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-chen26fk %I PMLR %P 17714--17733 %U https://proceedings.mlr.press/v306/chen26fk.html %V 306 %X Self-supervised learning for physiological time-series aims to captures the identity of the underlying dynamical process while filtering irrelevant noise. However, existing approaches may obscure the clinical semantics important for downstream transferability. Weakly constrained pretext tasks (i.e. contrastive learning, MAE) may incorrectly ignore the underlying dynamical structure, while structurally constrained models (i.e. SVAEs) are unable to selectively filter sample-specific noise. To bridge this gap, we propose ${\bf PULSE}$, a novel pretraining objective that simultaneously preserves dynamical relationships important to physiological time-series while selectively removing irrelevant noise. We achieve this by formulating a dynamical systems model to identify transferable and non-transferable information between time-series windows, and target the former through a novel cross-reconstruction objective. We establish theory that provides conditions for when transferrable information is recovered, and empirically validate it through synthetic experiments. On several real-world datasets, PULSE effectively distinguishes clinical semantic classes, increases label efficiency, and improves transfer learning performance.
APA
Chen, Y., Xu, M.A., Rehg, J.M. & Rozell, C.J.. (2026). Self-Supervised Dynamical System Representations for Physiological Time-Series. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:17714-17733 Available from https://proceedings.mlr.press/v306/chen26fk.html.

Related Material