Metraipolis › archive › dc_aa162923c1cd

When Does Online Adaptation Pay on the Edge? A Leakage-Free Evaluation of Warmup, Learning-Rate Selection, and Resource Trade-offs for Time-Series Forecasting — arXiv:2609.01126

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.01126
License (as archived)
CC0-1.0
Archived
2026-10-17T06:52:18.774Z
SHA-256 of the archived text
9b0ea87ac5608d67fdd79816d8ff3a9948749b98f6b8c7d6db5463c4e462776c
Stable URL of this page
https://metraipolis.com/documento/dc_aa162923c1cd

How to cite

When Does Online Adaptation Pay on the Edge? A Leakage-Free Evaluation of Warmup, Learning-Rate Selection, and Resource Trade-offs for Time-Series Forecasting — arXiv:2609.01126 — arXiv (en), https://arxiv.org/abs/2609.01126 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_aa162923c1cd (Metraipolis archive dc_aa162923c1cd, sha256 9b0ea87ac5608d67fdd79816d8ff3a9948749b98f6b8c7d6db5463c4e462776c)
BibTeX
@misc{dcaa162923c1cd,
  title = {When Does Online Adaptation Pay on the Edge? A Leakage-Free Evaluation of Warmup, Learning-Rate Selection, and Resource Trade-offs for Time-Series Forecasting — arXiv:2609.01126},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.01126},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_aa162923c1cd; sha256 9b0ea87ac5608d67fdd79816d8ff3a9948749b98f6b8c7d6db5463c4e462776c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.01126. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Online adaptation can help edge time-series forecasting under distribution drift, but its measured benefit is sensitive to evaluation choices. We study six public multivariate streams, including building-sensor and smart-meter data, under a leakage-free streaming protocol. We identify two additional sources of comparison bias. First, the warmup budget of the static baseline has a two-sided effect: insufficient warmup undertrains the baseline, whereas excessive warmup can degrade its pre-drift generalization. Across six dataset-backbone settings, the estimated adaptation benefit changes by 3.0 to 18.8 percentage points (pp) over the 1,000-20,000-step warmup range. Second, comparing SGD with momentum (SGD+m) and Adam at a shared default learning rate conflates optimizer quality with rate sensitivity. We select both the warmup budget and each optimizer's online rate using a held-out pre-drift validation slice without accessing test data. Under this validation-only procedure, Adam outperforms SGD+m in 310 of 360 evaluated cells, while 4 Adam cells remain below the static baseline. We further characterize accuracy against adaptation-state memory and A100-measured per-update latency for full, head-only, and calibration-based adaptation. In the evaluated PatchTST frontier settings, several parameter-efficient variants are nondominated on the adaptation-state-memory axis. Smart-meter analyses also show that reported gains depend on meter-selection rules. These findings support a validation-only commissioning procedure, while target-device latency and energy remain to be measured. Code, data, and all reported numbers: this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.01126 — 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 2026-10-17T06:41:55.840Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).