Browse by author
Lookup NU author(s): Dr Vicky FawcettORCiD
This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).
© 2025 The Author(s). The Dark Energy Spectroscopic Instrument (DESI) survey uses an automatic spectral classification pipeline to classify spectra. QuasarNET is a convolutional neural network used as part of this pipeline originally trained using data from the Baryon Oscillation Spectroscopic Survey (BOSS). In this paper we implement an active learning algorithm to optimally select spectra to use for training a new version of the QuasarNET weights file using only DESI data, with the goal of improving classification accuracy. This active learning algorithm includes a novel outlier rejection step using a Self-Organizing Map to ensure we label spectra representative of the larger quasar sample observed in DESI. We perform two iterations of the active learning pipeline, assembling a final dataset of 5600 labeled spectra, a small subset of the approximately 1.3 million quasar targets in DESI's Data Release 1. When splitting the spectra into training and validation subsets we achieve similar performance to the previously trained weights file in completeness and purity calculated on the validation dataset but do so with less than one tenth of the amount of training data. The new weights also more consistently classify objects in the same way when used on unlabeled data compared to the old weights file. In the process of improving QuasarNET's classification accuracy we discovered a systemic error in QuasarNET's redshift estimation and used our findings to improve our understanding of QuasarNET's redshifts.
Author(s): Green D, Kirkby D, Aguilar J, Ahlen S, Alexander DM, Armengaud E, Bailey S, Bault A, Bianchi D, Brodzeller A, Brooks D, Claybaugh T, de Belsunce R, de la Macorra A, Doel P, Fawcett VA, Ferraro S, Font-Ribera A, Forero-Romero JE, Gaztanaga E, Gontcho SA, Gutierrez G, Ishak M, Juneau S, Kehoe R, Kisner T, Kremin A, Lambert A, Landriau M, Le Guillou L, Levi ME, Manera M, Meisner A, Miquel R, Moustakas J, Myers AD, Palanque-Delabrouille N, Prada F, Perez-Rafols I, Rossi G, Sanchez E, Saulder C, Schlegel D, Schubnell M, Seo H, Sinigaglia F, Sprayberry D, Tan T, Tarle G, Weaver BA, Youles S, Yu J, Zhou R, Zou H
Publication type: Article
Publication status: Published
Journal: Journal of Cosmology and Astroparticle Physics
Year: 2025
Volume: 2025
Issue: 10
Online publication date: 23/10/2025
Acceptance date: 22/09/2025
Date deposited: 03/11/2025
ISSN (electronic): 1475-7516
Publisher: Institute of Physics
URL: https://doi.org/10.1088/1475-7516/2025/10/087
DOI: 10.1088/1475-7516/2025/10/087
Data Access Statement: All data and python code used to generate the plots in this paper are accessible at https://doi.org/10.5281/zenodo.15328537.
Altmetrics provided by Altmetric
