Real-Time Active Learning for optimised spectroscopic follow-up: Enhancing early SN Ia classification with the Fink broker
Abstract
Current and future surveys rely on machine learning classification to obtain large and complete samples of transients. Many of these algorithms are restricted by training samples that contain a limited number of spectroscopically confirmed events. Here, we present the first real-time application of Active Learning to optimise spectroscopic follow-up with the goal of improving training sets of early type Ia supernovae (SNe Ia) classifiers. Using a photometric classifier for early SN Ia, we apply an Active Learning strategy for follow-up optimisation using the real-time FINK broker processing of the ZTF public stream. We perform follow-up observations at the ANU 2.3m telescope in Australia and obtain 92 spectroscopic classified events that are incorporated in our training set. We show that our follow-up strategy yields a training set that, with 25% less spectra, improves classification metrics when compared to publicly reported spectra. Our strategy selects in average fainter events and, not only supernovae types, but also microlensing events and flaring stars which are usually not incorporated on training sets. Our results confirm the effectiveness of active learning strategies to construct optimal training samples for astronomical classifiers. With the Rubin Observatory LSST soon online, we propose improvements to obtain earlier candidates and optimise follow-up. This work paves the way to the deployment of real-time AL follow-up strategies in the era of large surveys.
Related
- Machine learning-based classification of type 1 active galactic nuclei
- Machine Learning Based Classification of Type 1 Active Galactic Nuclei
- Spectroscopic Study of the Circularly Polarised Radio Source TYC 2834-1385-1: An Active Binary
- Changing-look Active Galactic Nuclei from the Dark Energy Spectroscopic Instrument. VII. Testing Disk–Corona Diagnostics with eROSITA
- The Population of Type~1 Active Galactic Nuclei in Dark Energy Spectroscopic Instrument. I. Profile-dependent Measurements of the Low-redshift Sample
BibTeX
@article{moller2026realtime,
title = {{Real-Time Active Learning for optimised spectroscopic follow-up: Enhancing early SN Ia classification with the Fink broker}},
author = {A Möller and E.E.O Ishida and J Peloton and O. Vidal Velázquez and J Soon and B Martin and M Cluver and M Leoni and E Taylor},
journal = {Publications of the Astronomical Society of Australia},
year = {2026},
month = oct,
doi = {10.1017/pasa.2025.20},
eprint = {2502.19555},
archivePrefix = {arXiv},
url = {https://hal.science/hal-04985998v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Real-Time Active Learning for optimised spectroscopic follow-up: Enhancing early SN Ia classification with the Fink broker
AU - A Möller
AU - E.E.O Ishida
AU - J Peloton
AU - O. Vidal Velázquez
AU - J Soon
AU - B Martin
AU - M Cluver
AU - M Leoni
AU - E Taylor
PY - 2026
DA - 2026/10/05
JO - Publications of the Astronomical Society of Australia
DO - 10.1017/pasa.2025.20
UR - https://hal.science/hal-04985998v1
AB - Current and future surveys rely on machine learning classification to obtain large and complete samples of transients. Many of these algorithms are restricted by training samples that contain a limited number of spectroscopically confirmed events. Here, we present the first real-time application of Active Learning to optimise spectroscopic follow-up with the goal of improving training sets of early type Ia supernovae (SNe Ia) classifiers. Using a photometric classifier for early SN Ia, we apply an Active Learning strategy for follow-up optimisation using the real-time FINK broker processing of the ZTF public stream. We perform follow-up observations at the ANU 2.3m telescope in Australia and obtain 92 spectroscopic classified events that are incorporated in our training set. We show that our follow-up strategy yields a training set that, with 25% less spectra, improves classification metrics when compared to publicly reported spectra. Our strategy selects in average fainter events and, not only supernovae types, but also microlensing events and flaring stars which are usually not incorporated on training sets. Our results confirm the effectiveness of active learning strategies to construct optimal training samples for astronomical classifiers. With the Rubin Observatory LSST soon online, we propose improvements to obtain earlier candidates and optimise follow-up. This work paves the way to the deployment of real-time AL follow-up strategies in the era of large surveys.
SN - 1323-3580
SN - 1448-6083
ER -