Metraipolis › archive › dc_84b7efe38763
WinoTS: Wavelet-based Self-Distillation for Time Series Models — arXiv:2609.39337
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.39337
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T12:45:14.055Z
- SHA-256 of the archived text
f08eb1b531267e298a949eb9533b7ef0d9c8e3fe92d1c7589c8960ba8855d7c7- Stable URL of this page
https://metraipolis.com/documento/dc_84b7efe38763
How to cite
WinoTS: Wavelet-based Self-Distillation for Time Series Models — arXiv:2609.39337 — arXiv (en), https://arxiv.org/abs/2609.39337 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_84b7efe38763 (Metraipolis archive dc_84b7efe38763, sha256 f08eb1b531267e298a949eb9533b7ef0d9c8e3fe92d1c7589c8960ba8855d7c7)
BibTeX
@misc{dc84b7efe38763,
title = {WinoTS: Wavelet-based Self-Distillation for Time Series Models — arXiv:2609.39337},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.39337},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_84b7efe38763; sha256 f08eb1b531267e298a949eb9533b7ef0d9c8e3fe92d1c7589c8960ba8855d7c7},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.39337. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Self-supervised pre-training of time series models is currently dominated by next-token prediction and reconstruction objectives. In continuous-valued domains, these paradigms often waste model capacity on high-frequency, point-wise noise at the expense of learning invariant structure. While invariance-based self-distillation has proven highly effective in computer vision, its application to temporal data remains largely underexplored. Effectively adapting such methods to time series requires carefully designed augmentations: spatial operations like cropping can shift the timing of repeating cycles or distort the signal, while basic jittering may provide limited variation. We introduce Wavelet-based self-distillation for time series (WinoTS), an invariance-based pre-training paradigm designed specifically for temporal signals. At its core, WinoTS leverages time-frequency augmentations to construct multi-scale structural views without distorting underlying signal dynamics. Across extensive evaluations, WinoTS outperforms state-of-the-art baselines in long-term forecasting, cross-domain zero-shot transfer, and unsupervised anomaly detection. Notably, linear probing on frozen WinoTS representations frequently surpasses fully supervised models trained from scratch. Systematic ablations demonstrate that WinoTS is a flexible, architecture-agnostic framework yielding gains across time series backbones, and establish that time-frequency transformations provide a principled alternative to vision-style spatial augmentations. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.AI). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.39337 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-07T12:14:14.056Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).