Identifying galaxies, quasars, and stars with machine learning: A new catalogue of classifications for 111 million SDSS sources without spectra

Clarke, A. O.; Scaife, A.; Greenhalgh, R. J. S.; Griguta, V.

doi:10.1051/0004-6361/201936770

Cited by 69 publications

(63 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This adds to the complexity of model optimization. The results on faint end extrapolation are reported to have a high impact on the estimation reliability (e.g., Shu et al 2019;Clarke et al 2020;Logan & Fotopoulou 2020). We achieved satisfactory extrapolation results in r < 23.5, which is 1.5 magnitude larger than the SDSS limit.…”

Section: Limitations and Possible Improvementsmentioning

confidence: 54%

Photometric selection and redshifts for quasars in the Kilo-Degree Survey Data Release 4

Nakoneczny

Bilicki

Pollo

et al. 2021

A&A

View full text Add to dashboard Cite

We present a catalog of quasars with their corresponding redshifts derived from the photometric Kilo-Degree Survey (KiDS) Data Release 4. We achieved it by training machine learning (ML) models, using optical ugri and near-infrared ZYJHKs bands, on objects known from Sloan Digital Sky Survey (SDSS) spectroscopy. We define inference subsets from the 45 million objects of the KiDS photometric data limited to 9-band detections, based on a feature space built from magnitudes and their combinations. We show that projections of the high-dimensional feature space on two dimensions can be successfully used, instead of the standard color-color plots, to investigate the photometric estimations, compare them with spectroscopic data, and efficiently support the process of building a catalog. The model selection and fine-tuning employs two subsets of objects: those randomly selected and the faintest ones, which allowed us to properly fit the bias versus variance trade-off. We tested three ML models: random forest (RF), XGBoost (XGB), and artificial neural network (ANN). We find that XGB is the most robust and straightforward model for classification, while ANN performs the best for combined classification and redshift. The ANN inference results are tested using number counts, Gaia parallaxes, and other quasar catalogs that are external to the training set. Based on these tests, we derived the minimum classification probability for quasar candidates which provides the best purity versus completeness trade-off: p(QSOcand) > 0.9 for r < 22 and p(QSOcand) > 0.98 for 22 < r < 23.5. We find 158 000 quasar candidates in the safe inference subset (r < 22) and an additional 185 000 candidates in the reliable extrapolation regime (22 < r < 23.5). Test-data purity equals 97% and completeness is 94%; the latter drops by 3% in the extrapolation to data fainter by one magnitude than the training set. The photometric redshifts were derived with ANN and modeled with Gaussian uncertainties. The test-data redshift error (mean and scatter) equals 0.009 ± 0.12 in the safe subset and −0.0004 ± 0.19 in the extrapolation, averaged over a redshift range of 0.14 < z < 3.63 (first and 99th percentiles). Our success of the extrapolation challenges the way that models are optimized and applied at the faint data end. The resulting catalog is ready for cosmology and active galactic nucleus (AGN) studies.

show abstract

Section: Limitations and Possible Improvementsmentioning

confidence: 54%

Photometric selection and redshifts for quasars in the Kilo-Degree Survey Data Release 4

Nakoneczny

Bilicki

Pollo

et al. 2021

A&A

View full text Add to dashboard Cite

show abstract

“…The impact of the 5σ limits requested in W3 and W4 bands was investigated on the star/galaxy/quasar data set provided by Clarke et al (2020). 2 The magnitude selection cuts W3 < 11.32 and W4 < 8.0 3 were found to remove a very large fraction (∼98 per cent) of extragalactic sources from the sample but, unfortunately, also potential radio stars.…”

Section: Cross-matching With Supplementary Infrared Surveysmentioning

confidence: 99%

Evolutionary map of the Universe (EMU): Compact radio sources in the scorpio field towards the galactic plane

Riggi

Umana

Trigilio

et al. 2021

Monthly Notices of the Royal Astronomical Society

View full text Add to dashboard Cite

We present observations of a region of the Galactic plane taken during the Early Science Program of the Australian Square Kilometre Array Pathfinder (ASKAP). In this context, we observed the scorpio field at 912 MHz with an uncompleted array consisting of 15 commissioned antennas. The resulting map covers a square region of ∼40 deg2, centred on (l, b) = (343.5°, 0.75°), with a synthesized beam of 24 × 21 arcsec2 and a background rms noise of 150–200 μJy beam−1, increasing to 500–600 μJy beam−1 close to the Galactic plane. A total of 3963 radio sources were detected and characterized in the field using the caesar source finder. We obtained differential source counts in agreement with previously published data after correction for source extraction and characterization uncertainties, estimated from simulated data. The ASKAP positional and flux density scale accuracy were also investigated through comparison with previous surveys (MGPS, NVSS) and additional observations of the scorpio field, carried out with ATCA at 2.1 GHz and 10 arcsec spatial resolution. These allowed us to obtain a measurement of the spectral index for a subset of the catalogued sources and an estimated fraction of (at least) 8 per cent of resolved sources in the reported catalogue. We cross-matched our catalogued sources with different astronomical data bases to search for possible counterparts, finding ∼150 associations to known Galactic objects. Finally, we explored a multiparametric approach for classifying previously unreported Galactic sources based on their radio-infrared colours.

show abstract

“…Particularly for quasars, despite how important they are to a wide range of astronomy studies and research, their sample sizes are still in the relative minority class (Clarke et al, 2020). It is only through improved classification processes can significant increases be made in the sample sizes of quasars and other celestial objects, consequently enabling further progress to be made in research.…”

Section: Significance and Literature Reviewmentioning

confidence: 99%

“…With modern telescopes recording an increasingly large amount of astronomical data, the usage of machine learning models in the task of classification has become more and more significant and prevalent because of their accuracy and speed (Clarke et al, 2020). While reviews of previous studies with similar objectives of classifications found both supervised and unsupervised machine learning models to be capable of great performance with accuracy and other metrics measured at over 90%, the supervised models had generally higher accuracies in classification, and unsupervised models were shown to be more effective at detecting unknown objects (Viquar et al, 2018;Zhang et al, 2013).…”

Section: Significance and Literature Reviewmentioning

confidence: 99%

“…It has always been the pursuit of astronomers to study, analyze, and understand the universe with its countless celestial objects of endless variety. However, due to the celestial objects' differing distances, light-emitting intensities, and positions in space, they often appear as nearidentical smudges or dots of light when recorded using photometric data from telescopes, making it difficult to differentiate them from each other (Clarke et al, 2020). Therefore, the classification of celestial objects is very fundamental in astronomy for astronomers to be able to collect meaningful samples to critically study these celestial objects and uncover the complex and mysterious past, present, and future of the universe (Clarke et al, 2020).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Machine Learning Classification of Stars, Galaxies, and Quasars

Wu¹

2021

mijst

View full text Add to dashboard Cite

The objective of this study was to create a predictive model to classify stars, galaxies, and quasars, along with comparing different classification models to find the superior one. I hypothesized that it was possible to successfully train a machine learning model to classify stars, galaxies, and quasars using astronomical data provided by the Sloan Digital Sky Survey (SDSS). A multinomial logistic regression model has been trained and tested. It had an accuracy of 0.87, a weighted average precision, recall, and an f-1 score of 0.87, and a cross-validation accuracy score of 0.8664. The next model, a decision tree, had an accuracy of 0.99, weighted average precision, recall, and an f-1 score of 0.99, a cross-validation accuracy score of 0.99, and a cross-validation accuracy score of 0.9858. The decision tree model had significantly superior performance compared to the logistic regression model and was a good fit and accurate classifier for stars, galaxies, and quasars, proving my hypothesis to be correct. The model from this study could be used as a reliable classification tool for a wide variety of astronomical purposes to accelerate the expansion of the sample sizes of stars, galaxies, and especially quasars.

show abstract

Identifying galaxies, quasars, and stars with machine learning: A new catalogue of classifications for 111 million SDSS sources without spectra

Cited by 69 publications

References 52 publications

Photometric selection and redshifts for quasars in the Kilo-Degree Survey Data Release 4

Photometric selection and redshifts for quasars in the Kilo-Degree Survey Data Release 4

Evolutionary map of the Universe (EMU): Compact radio sources in the scorpio field towards the galactic plane

Machine Learning Classification of Stars, Galaxies, and Quasars

Contact Info

Product

Resources

About