aiqbench analysis
TS2TabPFN bridges the gap between manual feature engineering and end-to-end deep learning by integrating explicit features with TabPFN 2.5, setting new state-of-the-art benchmarks in time series classification and extrinsic regression. This approach demonstrates that combining foundation models for tabular data with structured temporal inputs effectively overcomes limitations inherent to single-paradigm architectures. The method offers a robust alternative for extracting value from ubiquitous time sequence data without relying solely on complex raw-data processing.
TS2TabPFN: Time Series Classification and Extrinsic Regression through Feature Extraction and a Tabular Foundation Model
arXiv:2608.04174v1 Announce Type: new Abstract: Time series data are ubiquitous in practical applications, where classification (TSC) and extrinsic regression (TSER) have emerged as essential tasks for obtaining value from temporal sequences. While the literature has seen significant progress through feature-based and deep learning models, existing methods often focus either on the quality of feature extraction or on the intrinsic predictive power of complex architectures applied to raw data. This division creates a gap between the control offered by feature engineering and the automated performance of end-to-end models. This paper proposes TS2TabPFN, a framework that bridges this gap by integrating explicit feature extraction with TabPFN 2.5, a cutting-edge foundation model for tabular data, to leverage its predictive capabilities. Our extensive experimental evaluation demonstrates that TS2TabPFN significantly outperforms state-of-the-art models in TSER tasks with statistical significance, providing a robust and efficient alternative for TSC and surpassing most of the currently best-performing algorithms. These results suggest that combining foundation models with structured features overcomes single-paradigm limitations, establishing a new time series state-of-the-art.
Abstract:Time series data are ubiquitous in practical applications, where classification (TSC) and extrinsic regression (TSER) have emerged as essential tasks for obtaining value from temporal sequences. While the literature has seen significant progress through feature-based and deep learning models, existing methods often focus either on the quality of feature extraction or on the intrinsic predictive power of complex architectures applied to raw data. This division creates a gap between the control offered by feature engineering and the automated performance of end-to-end models. This paper proposes TS2TabPFN, a framework that bridges this gap by integrating explicit feature extraction with TabPFN 2.5, a cutting-edge foundation model for tabular data, to leverage its predictive capabilities. Our extensive experimental evaluation demonstrates that TS2TabPFN significantly outperforms state-of-the-art models in TSER tasks with statistical significance, providing a robust and efficient alternative for TSC and surpassing most of the currently best-performing algorithms. These results suggest that combining foundation models with structured features overcomes single-paradigm limitations, establishing a new time series state-of-the-art.
Accepted at ECML PKDD 2026. To appear in Springer LNCS. 18 pages; 7 figures
Machine Learning (cs.LG)
arXiv:2608.04174 [cs.LG]
(or arXiv:2608.04174v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2608.04174 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
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 · 272 words
Related coverage
Fast, Interpretable, and Deterministic Time Series Classification With a Bag-of-Receptive-Fields
arXiv · Oct 7, 2026
Exploding variance of means of exponentials: least-squares to the rescue – Machine Learning Research Blog
francisbach.com · Sep 27, 2026
7 Machine Learning Algorithms That Still Matter
KDnuggets · Jul 30, 2026
SMart: A Multi-source Multi-phase Time Series Representation Transfer Framework
arXiv · Sep 3, 2026