aiqbench analysis
This research introduces PULSE, a self-supervised pretraining framework for physiological time series that leverages dynamical system structures to isolate shared system parameters from sample-specific noise. By theoretically grounding the extraction of generative variables, the method improves label efficiency and transfer learning capabilities across diverse real-world datasets. The approach offers a robust alternative to heuristic strategies, providing a principled way to enhance representation learning in medical signal processing.
Self-Supervised Dynamical System Representations for Physiological Time-Series
arXiv:2512.00239v2 Announce Type: replace Abstract: The effectiveness of self-supervised learning (SSL) for physiological time series depends on the ability of a pretraining objective to preserve information about the underlying physiological state while filtering out unrelated noise. However, existing strategies are limited due to reliance on heuristic principles or poorly constrained generative tasks. To address this limitation, we propose a pretraining framework that exploits the information structure of a dynamical systems generative model across multiple time-series. This framework reveals our key insight that class identity can be efficiently captured by extracting information about the generative variables related to the system parameters shared across similar time series samples, while noise unique to individual samples should be discarded. Building on this insight, we propose PULSE, a cross-reconstruction-based pretraining objective for physiological time series datasets that explicitly extracts system information while discarding non-transferrable sample-specific ones. We establish theory that provides sufficient conditions for the system information to be recovered, and empirically validate it using a synthetic dynamical systems experiment. Furthermore, we apply our method to diverse real-world datasets, demonstrating that PULSE learns representations that can broadly distinguish semantic classes, increase label efficiency, and improve transfer learning.
Abstract:The effectiveness of self-supervised learning (SSL) for physiological time series depends on the ability of a pretraining objective to preserve information about the underlying physiological state while filtering out unrelated noise. However, existing strategies are limited due to reliance on heuristic principles or poorly constrained generative tasks. To address this limitation, we propose a pretraining framework that exploits the information structure of a dynamical systems generative model across multiple time-series. This framework reveals our key insight that class identity can be efficiently captured by extracting information about the generative variables related to the system parameters shared across similar time series samples, while noise unique to individual samples should be discarded. Building on this insight, we propose PULSE, a cross-reconstruction-based pretraining objective for physiological time series datasets that explicitly extracts system information while discarding non-transferrable sample-specific ones. We establish theory that provides sufficient conditions for the system information to be recovered, and empirically validate it using a synthetic dynamical systems experiment. Furthermore, we apply our method to diverse real-world datasets, demonstrating that PULSE learns representations that can broadly distinguish semantic classes, increase label efficiency, and improve transfer learning.
Machine Learning (cs.LG); Machine Learning (stat.ML)
arXiv:2512.00239 [cs.LG]
(or arXiv:2512.00239v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2512.00239 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 282 words
Related coverage
Prototype-based Self-Supervised Multimodal Learning for PPG and Accelerometry Signals
arXiv · Aug 6, 2026
STAR: Adaptive Spatial-Temporal Normalization for Unified Microservice Incident Management
arXiv · Sep 30, 2026
Neonatal Hypoxic-ischaemic Encephalopathy Classification from the EEG and HRV Signals Using a Conformer based Masked Autoencoder
arXiv · Jul 28, 2026
READ-Bench: Benchmarking Historical Instance Retrieval for Time-Series Diagnosis
arXiv · Sep 29, 2026