Metraipolis › archive › dc_f971ec45b63d
TMAML: Temporal Model-Agnostic Meta-Learning for Cold-Start Time Series Forecasting — arXiv:2610.04547
- Source
- arXiv
- Original
- https://arxiv.org/abs/2610.04547
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-06T05:30:36.944Z
- SHA-256 of the archived text
80ede1594ca6d5626cd7695e3131bc1a8afc7d8df3a7415bd32f20cd9a018473- Stable URL of this page
https://metraipolis.com/documento/dc_f971ec45b63d
How to cite
TMAML: Temporal Model-Agnostic Meta-Learning for Cold-Start Time Series Forecasting — arXiv:2610.04547 — arXiv, https://arxiv.org/abs/2610.04547 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f971ec45b63d (Metraipolis archive dc_f971ec45b63d, sha256 80ede1594ca6d5626cd7695e3131bc1a8afc7d8df3a7415bd32f20cd9a018473)
BibTeX
@misc{dcf971ec45b63d,
title = {TMAML: Temporal Model-Agnostic Meta-Learning for Cold-Start Time Series Forecasting — arXiv:2610.04547},
howpublished = {arXiv},
url = {https://arxiv.org/abs/2610.04547},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f971ec45b63d; sha256 80ede1594ca6d5626cd7695e3131bc1a8afc7d8df3a7415bd32f20cd9a018473},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.04547. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Cold-start forecasting, the task of forecasting a time series with little to no historical data, is a common challenge. Addressing it requires approaches that learn quickly from few datapoints and leverage information from related series, typically through static covariates, to generalize well to unseen series. While some global forecasting models can generate cold-start predictions by leveraging information shared across multiple series, they are not optimized for out-of-train-set generalization or adaptation from short histories. In this work, we formulate cold-start forecasting as a few-window learning problem and introduce Temporal Model-Agnostic Meta-Learning (TMAML), which tailors the model-agnostic meta-learning algorithm, originally developed for few-shot adaptation of neural networks, to deep time series forecasting. TMAML constructs meta-tasks as temporally consistent support-query windows and pairs them with a temporal meta-training and meta-testing procedure, yielding forecasting models that are explicitly optimized for cold-start forecasting. We instantiate TMAML on the Temporal Fusion Transformer (TFT) and present an initial empirical analysis of forecast accuracy and calibration across three cold-start scenarios: TMAML consistently outperforms or matches a standard ERM-trained TFT, yields better-calibrated forecasts than naive on two of the three scenarios, but does not consistently outperform naive on probabilistic forecast accuracy. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2610.04547 — 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 2026-10-06T05:27:05.690Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).