EUVIMEDEUVIMED entdecken
Crossref

A multi-task deep learning approach for automatic classification and redshift measurement of astronomical spectra

Mingyu Zheng, Liangping Tu, Zhenyang Xu, Yue Jiang, Yuanjin Wang, Yu Mao, Yuanbao Lu

Publications of the Astronomical Society of Japan · 04.06.2026 · 78 (3) · 958-975

DOI: 10.1093/pasj/psag032 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Publications of the Astronomical Society of Japan, online erschienen am 24.03.2026. Verfasst von Mingyu Zheng und 6 weiteren Autor:innen. Die Arbeit nennt 37 Literaturangaben. Gefördert von: National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

To address the demand for processing massive astronomical spectral data generated by large-scale sky survey projects, as well as the issues of existing methods—such as ambiguous classification boundaries in feature-overlapping scenarios and insufficient measurement accuracy under low SNR or high redshift conditions—this study proposes SpecCZ-Net, a deep learning model integrating multi-module collaborative feature extraction. Based on a 1D-CNN, the model incorporates Gabor filters and the SE attention mechanism and realizes the simultaneous output of celestial object types and redshift values through a main classification branch and a multi-sub-branch parallel structure. To verify the model performance, two datasets were constructed: a simulated dataset generated based on galaxy templates (covering four types: S0, Sa, Sb, Sc) and an empirical dataset from the SDSS DR17 (including five categories: Star, Galaxy_AGN, Galaxy_STARBURST, Galaxy_STARFORMING, and QSO_BROADLINE). The results demonstrate that on the empirical dataset, SpecCZ-Net achieves an average classification accuracy of 95.40%—with the F1-score of the QSO_BROADLINE category reaching as high as 99.40%, and those of Galaxy_AGN and Galaxy_STARBURST being 94.28% and 94.01%, respectively. In terms of redshift measurement, $R^2$ can reach up to 99.83%, the NMAE is as low as 0.0015, and the GF remains stable within the range of 93% to 95%. Specifically, for QSO_BROADLINE, the NMAE is 0.0169 and the GF is 93.03%. Overall, its performance is comprehensively superior to that of the other five comparative baseline models. Five independent repeated experiments confirm the model’s excellent stability and reproducibility. Finally, model visualization analysis verifies that its attention weight distribution is highly consistent with the positions of characteristic spectral lines (e.g., [O ii], [O iii], Mg ii) provided by SDSS for the same spectrum. SpecCZ-Net provides a new approach for the automated and high-precision processing of massive astronomical spectral data.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Zheng, M., Tu, L., Xu, Z., Jiang, Y., Wang, Y., Mao, Y., & Lu, Y. (2026). A multi-task deep learning approach for automatic classification and redshift measurement of astronomical spectra. Publications of the Astronomical Society of Japan, 78(3), 958-975. https://doi.org/10.1093/pasj/psag032

Vancouver

Zheng M, Tu L, Xu Z, Jiang Y, Wang Y, Mao Y, et al. A multi-task deep learning approach for automatic classification and redshift measurement of astronomical spectra. Publications of the Astronomical Society of Japan. 2026;78(3):958-975. doi:10.1093/pasj/psag032

BibTeX
@article{Zheng2026,
  author = {Zheng, Mingyu and Tu, Liangping and Xu, Zhenyang and Jiang, Yue and Wang, Yuanjin and Mao, Yu and Lu, Yuanbao},
  title = {A multi-task deep learning approach for automatic classification and redshift measurement of astronomical spectra},
  journal = {Publications of the Astronomical Society of Japan},
  year = {2026},
  volume = {78},
  number = {3},
  pages = {958--975},
  doi = {10.1093/pasj/psag032}
}
Kostenlos starten →