TSPFN: A Temporal Tabular Foundation Model for Physiological Time Series Classification
2026-08-31 • Machine Learning
Machine LearningComputer Vision and Pattern Recognition
AI summaryⓘ
The authors developed TSPFN, a new machine learning model designed to better understand medical time-series data, like heart rate signals. Unlike previous models that treat data like simple tables, TSPFN understands how data changes over time by including special time-related information. They trained it on a large set of real medical data, which helped it perform better than existing models across different medical tasks. Their work shows that considering time details improves how well models learn from limited medical data.
physiological time seriesmachine learningfoundation modelsTabPFNpositional embeddingsspatio-temporal representationpretrainingcross-domain generalizationmedical time-series classificationin-context learning
Authors
Jérémie Stym-Popper, Clément Rambour, Federica Granese, Nicolas Thome, Olivier Bernard
Abstract
Designing models that generalize effectively in low- to medium-data regimes remains a primary challenge in medical machine learning, particularly for physiological time-series classification. While tabular foundation models such as TabPFN offer an attractive alternative to conventional fine-tuning through in-context learning, they are not designed to capture the temporal dependencies inherent to physiological signals. ~In this paper, we introduce TSPFN, a foundation model that redesigns TabPFN's architecture for time series data. TSPFN integrates structured temporal representations and positional embeddings to capture intra-sample temporal and channel dependencies. To fully leverage its spatio-temporal design, the model is pretrained on 140,000 real-world physiological time series across multiple medical domains. This yields a unified, generalizable framework capable of learning the specificities of medical time series. Experiments across diverse physiological benchmarks demonstrate that TSPFN consistently outperforms standard tabular baselines and TabPFN, and achieves superior cross-domain generalization compared to specialized deep time-series models. All our experiments, ablation studies, and pre-processing scheme are publicly available at https://github.com/Jeremstym/TSPFN