Metraipolis › archive › dc_767963cfd67f

Does High Signal-to-Noise Ratio Identify Better Training Data for Seismic Deep Learning? — arXiv:2609.37463

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.37463
License (as archived)
CC0-1.0
Archived
2026-10-07T20:15:01.931Z
SHA-256 of the archived text
d3032dac812f382f299ca15d6c0ab549698de947ae0ed633698e93b3238840fe
Stable URL of this page
https://metraipolis.com/documento/dc_767963cfd67f

How to cite

Does High Signal-to-Noise Ratio Identify Better Training Data for Seismic Deep Learning? — arXiv:2609.37463 — arXiv (en), https://arxiv.org/abs/2609.37463 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_767963cfd67f (Metraipolis archive dc_767963cfd67f, sha256 d3032dac812f382f299ca15d6c0ab549698de947ae0ed633698e93b3238840fe)
BibTeX
@misc{dc767963cfd67f,
  title = {Does High Signal-to-Noise Ratio Identify Better Training Data for Seismic Deep Learning? — arXiv:2609.37463},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.37463},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_767963cfd67f; sha256 d3032dac812f382f299ca15d6c0ab549698de947ae0ed633698e93b3238840fe},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.37463. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Hard signal-to-noise ratio (SNR) screening removes weak seismic records, but does retaining clearer waveforms improve learning enough to compensate for the discarded data? We test this question in phase picking and ambient-noise dispersion, first matching training counts and then restoring the full eligible dispersion pool at fixed compute. Phase-picking screening gave no consistent unfiltered-test benefit under fine-tuning or random initialization, including a control matched on distance and phase composition. Dispersion screening increased error at the original five-epoch endpoint but improved on an equally sized unscreened set after longer training. Retaining all 29,788 eligible fit paths instead of 9,929 randomly sampled paths reduced unfiltered-test error by 3.0% across three seeds at the same update budget. Full-available and strict-SNR training then had similar mean errors, 0.0405 and 0.0406 km s$^{-1}$, with mixed seed-paired ordering; strict training retained a 9.9% advantage on the SNR-selected test. In two-day monitoring, SNR regulated pick streams but event recovery depended on its measurement definition. High SNR did not reliably identify better training data: gains depended on the task, training schedule, and test domain, and fixed-count comparisons omitted the benefit of retaining additional valid examples.

Tema degli esploratori: terra, clima ed energia (retro-catalogo arXiv: annunci del 2026-09-30 in physics.geo-ph).
Categorie arXiv: Geophysics (physics.geo-ph).
Pagina del lavoro: https://arxiv.org/abs/2609.37463 — 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 2026-10-07T19:34:47.291Z dalla pagina https://arxiv.org/catchup/physics.geo-ph/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).