The Next Generation Fornax Survey (NGFS).VIII. A Support Vector Machine Approach for Disentangling Globular Clusters from other Sources
Yasna Ordenes-Briceño, Thomas H. Puzia, Paul Eigenthaler, Matias Blaña, Juan P. Carvajal, Matthew A. Taylor, Bryan W. Miller, Rohan Rahatgaonkar, Evelyn J. Johnston, Prasanta K. Nayak, Gaspar Galaz · 2025
Read The Next Generation Fornax Survey (NGFS).VIII. A Support Vector Machine Approach for Disentangling Globular Clusters from other Sources by Yasna Ordenes-Briceño, Thomas H. Puzia, Paul Eigenthaler, Matias Blaña, Juan P. Carvajal, Matthew A. Taylor, Bryan W. Miller, Rohan Rahatgaonkar, Evelyn J. Johnston, Prasanta K. Nayak, Gaspar Galaz (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-15895.txt.
Introduction
Yasna Ordenes-Briceno and co authors use a support vector machine to separate globular clusters from stars and background galaxies in the Fornax cluster in a 2025 paper. Wide field surveys detect millions of unresolved sources, and clean globular cluster samples are needed to trace galaxy assembly. Manual sorting cannot scale to coming survey volumes.
Using the central 3 square degrees of the Next Generation Fornax Survey, the team trains on spectroscopically confirmed sources with 15 features from colors and basic shape measures. The full model reaches 97.3 percent accuracy, and a pruned 7 feature model reaches 96.6 percent. Dropping near ultraviolet or near infrared bands hurts performance, a lesson for LSST, Euclid, and Roman survey planning.
Catalog record
| CREATOR | Yasna Ordenes-Briceño, Thomas H. Puzia, Paul Eigenthaler, Matias Blaña, Juan P. Carvajal, Matthew A. Taylor, Bryan W. Miller, Rohan Rahatgaonkar, Evelyn J. Johnston, Prasanta K. Nayak, Gaspar Galaz |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 87 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-15895
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The Next Generation Fornax Survey (NGFS).VIII. A Support Vector Machine Approach for Disentangling Globular Clusters from other Sources Yasna Ordenes-Briceño, Thomas H. Puzia, Paul Eigenthaler, Matias Blaña, Juan P. Carvajal, Matthew A. Taylor, Bryan W. Miller, Rohan Rahatgaonkar, Evelyn J. Johnston, Prasanta K. Nayak, Gaspar Galaz Abstract Context. Wide‑field, multi‑band surveys now detect millions of unresolved sources in nearby galaxy clusters, yet separating globular clusters (GCs) from foreground stars and background galaxies remains challenging. Scalable, automated classification is therefore essential to convert the forthcoming data from facilities such as the Vera C. Rubin Observatory’s Legacy Survey of Space and Time (LSST) and the Nancy Grace Roman and Euclid space telescopes into robust constraints on galaxy assembly and dark‑matter halos. Aims. We introduce a supervised classification method to separate GCs, stars, and galaxies based on their locations in color–color diagrams. The main objective is to recover a clean GC sample for future scientific analysis. The method exploits broad spectral energy distribution (SED) coverage, deep photometry, and is optimized for next-generation survey volumes. Methods. We use the central 3 deg 2 of the Next Generation Fornax Survey (NGFS), which images the Fornax cluster in u ′ g ′ i ′ u^{\prime}g^{\prime}i^{\prime} (BLANCO/DECam) and J K s JK_{s} (VISTA/VIRCAM). We build a Support Vector Machine (SVM; svm.SVC , scikit-learn ) using 15 features: all color combinations of u ′ g ′ i ′ J K s u^{\prime}g^{\prime}i^{\prime}JK_{s} and basic morphological parameters (e.g. FWHM, ellipticity). Spectroscopically confirmed sources define the training classes. Results. Color pairs connecting near-UV/optical to near-IR - particularly ( u ′ − g ′ u^{\prime}\!-\!g^{\prime} ) vs. ( g ′ − K s g^{\prime}\!-\!K_{s} ) - yield the strongest discrimination between classes. The full 15 feature model achieves 97.3% accuracy and a pruned 7 feature model built from the most informative, least correlated features achieves 96.6% accuracy. Misclassifications amount 8.4% and 10.4%, respectively. Omitting the u ′ u^{\prime} or/and near-IR bands degrades performance. Emulating LSST filters with NGFS u ′ g ′ i ′ u^{\prime}g^{\prime}i^{\prime} and Dark Energy Survey (DES) r ′ z ′ Y r^{\prime}z^{\prime}Y shows that u ′ u^{\prime} and Y Y bands are crucial, but models lacking NIR remain suboptimal. The final 7 feature u ′ g ′ i ′ J K s u^{\prime}g^{\prime}i^{\prime}JK_{s} classifier will underpin forthcoming NGFS GC studies. Conclusions. Combining broad SED coverage with simple morphological parameters enables precise, scalable separation of unresolved sources. Including NIR bands significantly improves GC classification, and joining LSST with forthcoming Euclid and Roman data will further enhance machine-learning frameworks for extragalactic surveys. † † orcid: 0000-0001-7966-7606 † † orcid: 0000-0003-0350-7061 † † orcid: 0000-0001-8654-0101 † † orcid: 0000-0003-2139-0944 † † orcid: 0000-0001-6584-7104 † † orcid: 0000-0003-3009-4928 † † orcid: 0000-0002-5665-376X † † orcid: 0000-0002-5350-0282 † † orcid: 0000-0002-2368-6469 † † orcid: 0000-0002-4638-1035 † † orcid: 0000-0002-8835-0739 1 Introduction Globular clusters (GCs) occupy a singular niche in astrophysics: they are among the oldest and simplest stellar systems in the Universe ( 103 ; 108 , e.g.) . They encode the earliest star formation and assembly events of their host galaxies ( 5 ; 17 ; 22 , e.g.) . Systematic studies over the past decades have shown that essentially every galaxy more massive than \sim\!10^{8}\,\mathrm{M_{\odot}} (stellar mass) hosts a GC system whose chemical compositions, kinematics, sizes, and spatial distribution mirror the host’s evolution and merger history ( 42 ; 85 ; 86 ; 79 ; 23 , e.g.) . Because GC colors, age - metallicity scaling relations, and specific frequencies correlate tightly with host halo mass ( 45 ; 50 ; 43 , e.g.) , they have become widely used tracers of dark matter assembly on galactic and cluster scales ( 25 ; 22 ) . In dense environments such as galaxy clusters, GCs also contribute to understanding environmental effects on galaxy evolution ( 95 ; 62 , e.g.) and the intracluster light (ICL) formation ( 78 ; 2 ; 66 ; 60 , e.g.) . Accurate identification of GCs serves as an additional means to comprehend the formation and evolution of galaxies. Deep imaging from state-of-the-art ground-based and space-based observatories has significantly improved the photometric characterization of GCs. Despite the depth and quality of modern imaging, separating GCs from stars and background galaxies remains challenging due to overlapping photometric and morphological properties ( 87 , e.g.) . Traditional selection criteria based on color cuts or structural parameters fail to disentangle these various stellar systems properly, producing contamination fractions of 30 – 70\, % in purely optical photometric samples ( 84 , e.g.) , unless sample statistics is overwhelmingly in favor of GCs, e.g. in brightest cluster galaxies ( 51 ) . Decontamination becomes particularly challenging in the central regions of galaxy clusters or in the vicinity of their massive and regular galaxies, where their light profile affects the detection and photometry of stars (foreground), GCs (cluster) and galaxies (background), which can significantly reduced a GC catalog purity, complicating the interpretation of GC samples ( 31 ; 63 , e.g.) . 74 have shown that by incorporating a broader wavelength baseline, particularly including the u^{\prime} and K_{s} -band, the separation between GCs and other sources becomes clearer, as these bands enhance the contrast in the spectral energy distribution (SED) properties between the stellar systems. The scientific return from extragalactic GC studies has grown in lock‐step with advances in wide‐field, optical/near‐IR imaging ( 74 ; 99 , e.g.) . Surveys such as the Next Generation Virgo Survey ( 38 , NGVS,) , the Next Generation Fornax Survey ( 73 ; 32 ; 76 , NGFS,) , PHANGS– HST ( 67 , e.g.) , and, more recently, deep Euclid pilot programs now detect thousands of GC candidates per pointing ( 88 ) , probing to larger galactocentric radii and lower surface‐brightness regimes than ever before. Spectroscopic studies are infeasible for the millions of candidates predicted by modern cosmological simulations ( 80 , e.g. E–MOSAICS, see) and for the 350,000 estimated GCs in the Euclid footprint ( 34 ) and \gtrsim 4\times 10^{6} GCs forecast to be visible in the Vera C. Rubin–LSST imaging ( 56 ; 101 ) . These data volumes demand scalable, fully automated classification pipelines. Machine-learning (ML) methods have already demonstrated superior performance over traditional approaches in many domains of astronomy ( 4 ; 89 ; 9 ; 23 ; 100 , e.g.) . Among ”classical” algorithms, the support vector machine (SVM) remains attractive because it yields interpretable decision boundaries ( 26 ; 81 ; 27 ) , is robust against the ”curse of dimensionality” ( 57 ; 49 , e.g.) , and can be tuned efficiently with modest training sets ( 20 ) . SVMs have been applied successfully for classification to galaxy morphology ( 54 ; 104 , e.g.) , stellar objects and transients ( 61 ) , spectral line classification ( 93 ) and for identifying active galactic nuclei (AGNs), galaxies, and stars, with \lesssim\!5\, % cross-contamination ( 68 ; 19 ; 106 ) . In this work we advance these efforts by developing a supervised SVM pipeline using the NGFS’s ultra-deep, broad spectral baseline imaging - spanning u^{\prime}g^{\prime}i^{\prime}JK_{s} and morphological