A Machine Learning Quasar Candidate Catalog with Photometric Redshifts from KiDS DR5 and AllWISE
- 1Chinese Academy of Sciences
- 2University of Southern California
Abstract
We present two complementary machine-readable catalogs for quasar (QSO) candidate selection in the Kilo-Degree Survey (KiDS) DR5. The KiDS-only catalog covers the full survey source sample and provides optical/near-infrared-based classifications together with QSO photometric redshift estimates derived without Wide-field Infrared Survey Explorer (WISE) information. The KiDS–AllWISE catalog is restricted to sources with AllWISE counterparts and provides WISE-assisted classifications and photometric redshifts. Separate stacking ensembles are trained to distinguish galaxies, stars, and QSOs. On a stratified held-out test set, the WISE-assisted classifier achieves an overall accuracy of 98.34%, a macroaveraged area under the receiver operating characteristic curve (ROC AUC) of 99.78%, and a QSO F1-score of 97.09%. Spatially disjoint R.A.-block validation yields QSO F1-scores of 95.87%–96.31%, indicating only modest spatial degradation. Feature-ablation tests show that W1/W2-based mid-infrared information provides the largest gain in QSO recovery, whereas Bayesian photometric redshift (BPZ)-derived quantities have only a minor effect. For QSO photometric redshift estimation, we train a LightGBM–XGBoost ensemble on Sloan Digital Sky Survey (SDSS) DR16 Quasar Catalog (DR16Q) quasars with valid optical and WISE photometry. On 94,057 held-out quasars, the WISE-assisted model achieves σ NMAD = 0.0633 and a catastrophic-outlier fraction of η = 12.0%; excluding WISE features degrades these values to σ NMAD = 0.0948 and η = 26.8%. Both catalogs provide probability-ranked QSO candidates and thresholded p Q samples for different follow-up priorities. As the candidates are not spectroscopically confirmed, statistical applications require independent calibration of the corresponding selection functions.
Related
- Cosmic-neighbor-associated distances to VHE and candidate VHE BL Lacs via SDSS photometric redshifts
- FOMO-Z. Optimal galaxy classification, photometric redshifts, and reconstruction of population distributions
- 2Flow2SN: Dusty Supernova Cosmology with Photometric Redshifts and Contamination using Neural Simulation-Based Inference
- Core-dominated non-Gaussian photometric redshifts in the COSMOS field: a normality test analysis
- Data-driven Galaxy Population Prior for Photometric Redshifts
BibTeX
@article{pang2026machine,
title = {{A Machine Learning Quasar Candidate Catalog with Photometric Redshifts from KiDS DR5 and AllWISE}},
author = {Sen Pang and Yanxia Zhang and Jinghang Shi},
journal = {The Astrophysical Journal Supplement Series},
year = {2026},
month = oct,
doi = {10.3847/1538-4365/aea41d},
url = {https://iopscience.iop.org/article/10.3847/1538-4365/aea41d},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - A Machine Learning Quasar Candidate Catalog with Photometric Redshifts from KiDS DR5 and AllWISE
AU - Sen Pang
AU - Yanxia Zhang
AU - Jinghang Shi
PY - 2026
DA - 2026/10/19
JO - The Astrophysical Journal Supplement Series
DO - 10.3847/1538-4365/aea41d
UR - https://iopscience.iop.org/article/10.3847/1538-4365/aea41d
AB - We present two complementary machine-readable catalogs for quasar (QSO) candidate selection in the Kilo-Degree Survey (KiDS) DR5. The KiDS-only catalog covers the full survey source sample and provides optical/near-infrared-based classifications together with QSO photometric redshift estimates derived without Wide-field Infrared Survey Explorer (WISE) information. The KiDS–AllWISE catalog is restricted to sources with AllWISE counterparts and provides WISE-assisted classifications and photometric redshifts. Separate stacking ensembles are trained to distinguish galaxies, stars, and QSOs. On a stratified held-out test set, the WISE-assisted classifier achieves an overall accuracy of 98.34%, a macroaveraged area under the receiver operating characteristic curve (ROC AUC) of 99.78%, and a QSO F1-score of 97.09%. Spatially disjoint R.A.-block validation yields QSO F1-scores of 95.87%–96.31%, indicating only modest spatial degradation. Feature-ablation tests show that W1/W2-based mid-infrared information provides the largest gain in QSO recovery, whereas Bayesian photometric redshift (BPZ)-derived quantities have only a minor effect. For QSO photometric redshift estimation, we train a LightGBM–XGBoost ensemble on Sloan Digital Sky Survey (SDSS) DR16 Quasar Catalog (DR16Q) quasars with valid optical and WISE photometry. On 94,057 held-out quasars, the WISE-assisted model achieves σ NMAD = 0.0633 and a catastrophic-outlier fraction of η = 12.0%; excluding WISE features degrades these values to σ NMAD = 0.0948 and η = 26.8%. Both catalogs provide probability-ranked QSO candidates and thresholded p Q samples for different follow-up priorities. As the candidates are not spectroscopically confirmed, statistical applications require independent calibration of the corresponding selection functions.
SN - 0067-0049
SN - 1538-4365
ER -