Time-series models remain reliable despite changes in data sampling rates
Quantifying the Sampling Frequency Invariance of Time-Series Foundation Models for Sensor Data Applications
Machine LearningArtificial Intelligence
Summary
When devices collect sensor data at different speeds, it can confuse computer models trained to understand that data. The authors checked how well four advanced time-series models keep working when the speed of data collection changes, as long as important information is still captured. They created new ways to measure how much the model's output and accuracy change with sampling speed. Their study helps people know when these models can be trusted even if sensor data comes at different rates.
What this means in practice
- •For sensor software engineers: Tune sensor data processing pipelines knowing when time-series models handle varying sampling speeds without loss of accuracy.
- •For iot device developers: Design devices that adjust data sampling rates for power saving while maintaining reliable model predictions using tested time-series models.
Authors
Sourish Gunesh Dhekane, Abinaya Kumar, Torsten Woertwein, Yan-Bo Lin, Thomas Ploetz, Shabnam Ghaffarzadegan
Abstract
Time-Series Foundation Models (TSFMs) are increasingly used in sensor applications, where sampling frequency (SF) varies across sensors, devices, and operating modes (e.g., power saving). Ideally, TSFMs must remain effective during such SF changes, provided the SF retains task-critical bandwidth. In this paper, we stress-test 4 prominent TSFMs: MOMENT, Chronos-2, Moirai-MoE, and TimesFM, to quantify their SF-invariance and investigate whether their behavior tracks the resulting loss of information bandwidth rather than the SF value itself. Our stress-test probes 3 dimensions: (i) the shift in embeddings, (ii) the degradation in forecasting performance, and (iii) the degradation in downstream performance, quantified by three novel metrics: FSI, FDI, and CDI, respectively. To our knowledge, this is the first study to place SF-invariance at the core of TSFM evaluations. Our results characterize: (i) how SF-invariance depends on the margin between signal's native-SF and the bandwidth its task requires, (ii) how much of that invariance the choice of TSFM explains, and (iii) how much of TSFM's performance degradation under a lower SF reflects information loss rather than a limitation of their abilities. These findings reframe SF-invariance as a measurable, signal-aware property of the (signal X TSFM) pair, and guide practitioners on when TSFMs display SF-invariance.