ASCL.net

Astrophysics Source Code Library

Making codes discoverable since 1999

ASCL Code Record

[submitted] Machine Learning-Based Supernova Classification with PR-AUC Optimization

This repository implements an optimized XGBoost-based framework for photometric classification of Type Ia supernovae, addressing class imbalance through PR-AUC and F1-score prioritization. The approach is designed for scalability in large-scale astronomical surveys such as LSST and ensures improved classification robustness compared to traditional metrics like ROC-AUC.

Code site:
https://github.com/mranuraggarg/supernovae_classification

Views: 12