EUVIMEDEUVIMED entdecken
Crossref

Compact and physically interpretable feature models for photometric type Ia supernova classification

Anurag Garg

Journal of Astrophysics and Astronomy · 07.08.2026 · 47 (2)

DOI: 10.1007/s12036-026-10175-y ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Astrophysics and Astronomy, online erschienen am 07.08.2026. Verfasst von Anurag Garg. Die Arbeit nennt 16 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Photometric classification of Type Ia supernovae is essential for modern time-domain surveys, where spectroscopic confirmation is not always feasible for the full transient sample. In this work, we investigate a compact and physically interpretable feature representation derived from multi-band light curves and evaluate its performance using gradient-boosted decision trees on the Supernova Photometric Classification Challenge (SPCC) dataset. The compact representation is derived from an initial pool of 31 light-curve features, reduced to 30 after removing redundant variables, and further optimized to a 16-feature model using systematic ablation and performance analysis. The final compact model achieves an F1–score of 0.844 on the held-out test set. This is consistent with k-fold cross-validation results (0.841 ± 0.006). The precision–recall area under the curve (PR-AUC) is 0.928, with similarly low variance across folds. Relative to our earlier 31-feature optimized XGBoost model for the same SPCC classification task (F1 ≈\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\approx $$\end{document} 0.923), the compact 16-feature representation retains strong classification performance (F1 = 0.844) while substantially improving interpretability and reducing feature-space complexity. The ablation results show that temporal evolution provides the dominant classification signal, while brightness, color, and variability features supply complementary information. A reduced core of approximately ten physically meaningful features retains a large fraction of the performance of the compact model, with only a small decrease in F1-score, indicating that reliable classification does not require large high-dimensional feature spaces. These results demonstrate that interpretable feature-based models can capture the essential astrophysical information needed for Type Ia photometric classification, with direct implications for survey cadence, filter coverage, and the design of transparent and efficient machine learning pipelines for future time-domain surveys.

Zitieren

APA

Garg, A. (2026). Compact and physically interpretable feature models for photometric type Ia supernova classification. Journal of Astrophysics and Astronomy, 47(2). https://doi.org/10.1007/s12036-026-10175-y

Vancouver

Garg A. Compact and physically interpretable feature models for photometric type Ia supernova classification. Journal of Astrophysics and Astronomy. 2026;47(2). doi:10.1007/s12036-026-10175-y

BibTeX
@article{Garg2026,
  author = {Garg, Anurag},
  title = {Compact and physically interpretable feature models for photometric type Ia supernova classification},
  journal = {Journal of Astrophysics and Astronomy},
  year = {2026},
  volume = {47},
  number = {2},
  doi = {10.1007/s12036-026-10175-y}
}
Kostenlos starten →