EUVIMEDEUVIMED entdecken
Crossref

Quantitative LIBS analysis of microalloyed steel using LASSO–LightGBM modeling with SHAP-based interpretability

Sheng Wang, Yun Tang, Penghui Zou, Enxin Guo, Zeen Yang, Jiale Liao, Xiang Zhang, Tengfeng Ni, Zhanke Peng

Journal of Analytical Atomic Spectrometry · 2025 · 40 (12) · 3563-3568

DOI: 10.1039/d5ja00314h ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Analytical Atomic Spectrometry, 2025. Verfasst von Sheng Wang und 8 weiteren Autor:innen. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 28 Literaturangaben. Gefördert von: Natural Science Foundation of Hunan Province.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The LASSO–LightGBM–SHAP framework integrates L 1 -regularized wavelength selection, LightGBM-based gradient boosting, and SHAP interpretation to achieve accurate and interpretable multi-element quantification from complex LIBS spectra.

Förderung

  • Natural Science Foundation of Hunan Province

Zitieren

APA

Wang, S., Tang, Y., Zou, P., Guo, E., Yang, Z., Liao, J., Zhang, X., Ni, T., & Peng, Z. (2025). Quantitative LIBS analysis of microalloyed steel using LASSO–LightGBM modeling with SHAP-based interpretability. Journal of Analytical Atomic Spectrometry, 40(12), 3563-3568. https://doi.org/10.1039/d5ja00314h

Vancouver

Wang S, Tang Y, Zou P, Guo E, Yang Z, Liao J, et al. Quantitative LIBS analysis of microalloyed steel using LASSO–LightGBM modeling with SHAP-based interpretability. Journal of Analytical Atomic Spectrometry. 2025;40(12):3563-3568. doi:10.1039/d5ja00314h

BibTeX
@article{Wang2025,
  author = {Wang, Sheng and Tang, Yun and Zou, Penghui and Guo, Enxin and Yang, Zeen and Liao, Jiale and Zhang, Xiang and Ni, Tengfeng and Peng, Zhanke},
  title = {Quantitative LIBS analysis of microalloyed steel using LASSO–LightGBM modeling with SHAP-based interpretability},
  journal = {Journal of Analytical Atomic Spectrometry},
  year = {2025},
  volume = {40},
  number = {12},
  pages = {3563--3568},
  doi = {10.1039/d5ja00314h}
}
Kostenlos starten →