Metraipolis › archive › dc_567abaebb9f1
Rolling Conformal Prediction in Sequential Model Training — arXiv:2609.26951
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.26951
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T10:57:48.766Z
- SHA-256 of the archived text
34dc3388f759fc2a151b01082c9aa053fbcfbfc0ce58fab494402df30c8d426e- Stable URL of this page
https://metraipolis.com/documento/dc_567abaebb9f1
How to cite
Rolling Conformal Prediction in Sequential Model Training — arXiv:2609.26951 — arXiv (en), https://arxiv.org/abs/2609.26951 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_567abaebb9f1 (Metraipolis archive dc_567abaebb9f1, sha256 34dc3388f759fc2a151b01082c9aa053fbcfbfc0ce58fab494402df30c8d426e)
BibTeX
@misc{dc567abaebb9f1,
title = {Rolling Conformal Prediction in Sequential Model Training — arXiv:2609.26951},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.26951},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_567abaebb9f1; sha256 34dc3388f759fc2a151b01082c9aa053fbcfbfc0ce58fab494402df30c8d426e},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.26951. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We introduce Rolling Conformal Prediction (rolling-CP), a distribution-free predictive inference method for the setting of sequential model training. Specifically, given a data stream $(X_1,Y_1),(X_2,Y_2),\dots$, at each time $n$ the trained model may depend on the observed history $\{(X_i,Y_i)\}_{i<n}$. This setting arises naturally in modern sequential training, including one-pass training over massive datasets and continual fine-tuning or test-time adaptation of language models during deployment. Rolling-CP first calibrates each incoming observation against the current predictor and then rolls it into future training. In this way, we avoid the need for data splitting. Remarkably, although the models at times $n=1,2,\dots$ may have entirely different properties and accuracy levels, for exchangeable data it is nonetheless possible to establish a guarantee of marginal coverage, with a familiar universal factor-two guarantee (a worst case guarantee of $1-2\alpha$ coverage, as compared to the target level $1-\alpha$), without any assumptions of stability or any restrictions on the model training process. For i.i.d. data streams, we further prove high-probability training-conditional validity uniformly over time; under stability conditions, coverage guarantees sharpen towards $1-\alpha$. Numerical experiments on sequential regression, multiclass SGD, and one-pass neural-network training further demonstrate the practical effectiveness of rolling-CP.
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-24 in cs.LG).
Categorie arXiv: Statistics Theory (math.ST) ; Machine Learning (cs.LG); Methodology (stat.ME); Machine Learning (stat.ML).
Pagina del lavoro: https://arxiv.org/abs/2609.26951 — 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 2026-10-08T10:06:16.793Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).