Pith. sign in

REVIEW 5 major objections 5 minor 48 references

Machine Learning Classification of Young Stellar Objects and Evolved Stars in the Magellanic Clouds Using the Probabilistic Random Forest Classifier

T0 review · 5 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

Pith's one-line read This paper claims that a Probabilistic Random Forest classifier, trained on 618 spectroscopically labeled sources with 12 photometric bands, separates five dusty stellar classes in the Magellanic Clouds with 89% overall accuracy, and that…

desk verdict Plausible 89% PRF recall on 618 SAGE dusty stars, but the missing evaluation details and thin novelty mean the headline is not yet established. read the letter →

arxiv 2504.14242 v1 pith:MQXXB2HO submitted 2025-04-19 astro-ph.GA

classification astro-ph.GA
keywords stellarclassificationyoungobjectsAGBstarsredsupergiantspost-AGBMagellanicCloudsprobabilisticrandomforestSMOTE
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper asks whether photometry alone can tell apart five kinds of dusty stars in the Magellanic Clouds: young stellar objects, carbon-rich AGB stars, oxygen-rich AGB stars, red supergiants, and post-AGB stars. Using 618 spectroscopically confirmed sources from the SAGE survey, it compares six machine-learning classifiers and reports that the Probabilistic Random Forest (PRF) wins with 89% overall accuracy on the test set. It also tests the Synthetic Minority Oversampling Technique (SMOTE) to balance the imbalanced classes, and finds that SMOTE leaves the overall accuracy unchanged, only shifting precision and recall for the oxygen-rich AGB and YSO classes. The claim matters because a reliable photometric classifier would let large surveys of the Magellanic Clouds be labeled without per-object spectroscopy.

What carries the argument

The central object is the Probabilistic Random Forest (PRF), introduced as an extension of the random forest that treats both features and labels as probability distributions and sends each data point probabilistically down tree branches, which is meant to absorb the noise and label uncertainty of real photometric catalogs. The other load-bearing component is SMOTE, an oversampling technique that creates synthetic minority-class samples by interpolating along line segments between a point and its nearest neighbours; the paper applies it to the training sets to balance five imbalanced classes. The feature space is the 12-band SAGE photometry (UMmag, BMmag, VMmag, IMmag, J2mag, H2mag, Ks2mag, IRAC1–IRAC4, and [24]), and the target is a five-way spectral class label from the SAGE spectral catalog.

What would settle it

Hold out a stratified 20% of the 618 spectroscopically confirmed sources, train PRF on the remainder with SMOTE applied only to the training fold, and measure recall on the held-out set; if the overall recall is substantially below 89% (e.g., below 80%) or the per-class confusion matrix changes, the central claim does not generalise. The paper gives no split ratio or random seed, so the experiment cannot currently be reproduced as reported.

Watch

Extended reading notes

Core claim

The central discovery is that a probabilistic variant of the random forest, which routes each object down decision trees using probability distributions rather than fixed splits, classifies dusty stellar sources in the Magellanic Clouds with 89% overall accuracy (recall-weighted) from 12 near- and mid-infrared photometric bands alone. In the paper's best model, carbon-rich AGB stars and post-AGB stars are recalled at 100%, red supergiants at 88%, young stellar objects at 88–92%, and oxygen-rich AGB stars at 64–73% depending on whether SMOTE is applied. SMOTE raises the precision of the minority oxygen-rich AGB and YSO classes but lowers their recall, so the weighted average stays at 0.89; the paper reports no change for the CAGB, PAGB, and RSG classes. The authors interpret this as PRF handling noisy photometric and label uncertainty better than random forest, k-nearest neighbours, support-vector machines, or Gaussian naive Bayes.

Load-bearing premise

The 89% accuracy rests on an unspecified train/test split and on SMOTE being applied only to the training partition; if the split is unrepresentative or synthetic samples leak into the test set, the reported score is inflated.

Editorial extensions

If this is right

  • If the 89% accuracy is reproducible, photometric surveys in the Magellanic Clouds can be pre-classified into the five dusty stellar types with high fidelity, reducing the need for expensive spectroscopy.
  • SMOTE does not improve the overall score, so for these data the limiting factor is not raw class imbalance but the intrinsic overlap between classes (notably OAGB vs. PAGB and RSG).
  • PRF's probabilistic routing makes it a suitable default classifier for noisy, low-S/N photometric surveys beyond the Magellanic Clouds.
  • The per-class recall values (100% for CAGB and PAGB, lower for OAGB) set a concrete baseline that future models or additional bands must beat.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • One consequence the paper leaves implicit is that the 89% is only as good as the evaluation protocol: no split ratio, cross-validation scheme, or random seed is specified, and SMOTE must be confined to the training fold for the number to be trustworthy.
  • The classifier's 50% precision for post-AGB stars suggests it over-predicts this rare class; a class-weighted loss or a two-stage 'rare-class first' classifier might do better than plain SMOTE.
  • Because the LMC and SMC differ in distance and metallicity, a single model over both clouds could hide systematic biases; testing the PRF on each cloud separately would show whether the accuracy is galaxy-dependent.
  • A natural extension is to feed variability features from time-series photometry into the same PRF, since many AGB stars are long-period variables; the probabilistic tree structure may handle missing epochs better than the fixed data matrix used here.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 5 minor

Summary. The paper applies six machine learning classifiers (Probabilistic Random Forest, Random Forest, K-Nearest Neighbor, SVC-poly, SVC-rbf, and Gaussian Naive Bayes) to a spectroscopically labeled sample of 618 dusty stellar objects in the Magellanic Clouds from the SAGE project, using 12 photometric bands. Five classes are considered: YSOs, CAGB, OAGB, RSG, and PAGB. The authors compare performance on the original imbalanced dataset ('Simple') and after applying SMOTE ('Augmented'). The central claim, stated in the abstract and Section 4, is that PRF achieves the highest overall accuracy of 89% based on the recall metric, and that SMOTE does not improve classification for CAGB, PAGB, and RSG while changing performance for OAGB and YSO classes. The paper is a short proceedings contribution and does not provide code, data, or a detailed evaluation protocol.

Significance. If the reported 89% recall figure were reproducible, the paper would offer a useful benchmark for photometric classification of dusty stars in the Magellanic Clouds, particularly because PRF is designed to handle noisy labels and the comparison spans six classifiers and two data treatments. The per-class metrics in Table 2 provide a compact summary that could guide future work. However, the scientific value is severely limited by the absence of a reproducible evaluation protocol: no train/test split ratio, no cross-validation scheme, no hyperparameters, no random seeds, and no support counts. These omissions mean the headline number cannot be independently verified or trusted as an unbiased estimate. The paper also does not provide code or data, so the empirical claims are not testable.

major comments (5)
  1. [Section 2 and Table 1] The train/test split ratio, stratification scheme, and random seed are not stated anywhere. The text says only that SMOTE was 'applied to the training datasets,' but it does not say what fraction of the 618 objects was held out, whether the split was stratified by class or by galaxy (LMC/SMC), or whether the test set was kept completely separate before SMOTE. Without this information, the reported 89% recall for PRF is not reproducible and its statistical meaning is unclear.
  2. [Section 2 and Table 1] The 'Augmented Data' column in Table 1 sums to 972, exceeding the original 618 objects. If SMOTE was applied to the full dataset before splitting, synthetic samples (and their nearest-neighbor originals) could appear in both training and test partitions, inflating the reported accuracy and making the test set unrepresentative of real SAGE measurements. The manuscript must specify the exact order of preprocessing and demonstrate that no synthetic sample or its original counterpart crosses the train/test boundary.
  3. [Table 2] Table 2 omits per-class support counts, so the printed weighted-average F1-scores cannot be verified. Using the original class totals from Table 1 (CAGB 174, OAGB 107, PAGB 37, RSG 94, YSO 206) as approximate supports, the Simple PRF per-class F1 values (0.97, 0.76, 0.67, 0.82, 0.91) yield a weighted F1 of approximately 0.87, not the reported 0.89. This discrepancy may be due to an unstated test-set composition, but as written the headline number is not derivable from the manuscript, and the reported 'accuracy 0.89' cannot be distinguished from a weighted average without support counts.
  4. [Abstract and Section 4] The paper conflates accuracy and recall. The abstract says '89% based on the recall metric,' while Table 2 lists an 'accuracy' row of 0.89 alongside a weighted-average recall of 0.89. In Section 4, recall values of 100% and 88% are called 'accuracy.' The authors should state clearly which metric is the headline, report the actual number of test examples per class, and explain why accuracy equals weighted recall in this setting, if it does.
  5. [Section 3] No hyperparameters are provided for any classifier: number of trees and tree depth for RF/PRF, number of neighbors for KNN, kernel parameters (C, gamma) for SVC, or the SMOTE settings (number of nearest neighbors, sampling strategy). Without these, the classifier comparison is not reproducible and the claim that PRF outperformed the others is not auditable. A concise table of hyperparameters and a statement of whether they were tuned (and with what procedure) is required.
minor comments (5)
  1. [References] The reference 'Chawla et al., 2011' cites the arXiv e-print (arXiv:1106.1813) rather than the original 2002 JAIR publication of SMOTE; the original should be cited.
  2. [Figures 1 and 2] Figure 2's confusion matrices should include axis labels (predicted vs. true class) and class labels, and Figure 1 should have error bars or at least point markers to distinguish classifiers and data treatments. The current captions are too terse to interpret the plots.
  3. [Section 4] The statement that SMOTE 'did not impact the classification accuracy for the CAGB, PAGB, and RSG categories' is based on recall only: Table 2 shows CAGB F1 falls from 0.97 to 0.92 and RSG F1 from 0.82 to 0.78 under SMOTE. The metric being referenced should be stated explicitly to avoid misleading the reader.
  4. [Table 2] The 'accuracy' row in a per-class classification report is not a class-wise metric; it should be reported separately, along with the number of test samples, so that the reader can interpret the weighted averages.
  5. [General] A data and code availability statement is missing. Even for a proceedings paper, stating whether the SAGE data products and any analysis scripts are publicly accessible would improve the manuscript's utility.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the 89% headline is an empirical accuracy measured on spectroscopically labeled SAGE data, not a quantity derived from the method's own assumptions.

full rationale

The paper makes no derivation claim that could collapse into its inputs. Its central result, stated in the Abstract and Section 4, is that 'the PRF classifier demonstrated the highest total accuracy, achieving 89%' based on recall, with per-class precision, recall, and F1 shown in Table 2. This is an empirical evaluation of supervised classifiers on spectroscopically labeled SAGE photometry, where the labels are external spectroscopic classifications and the features are 12 observed filters. The PRF method is introduced through independent methodological citations (Baron 2019; Reis et al. 2019; Kinson et al. 2021, 2022), and the self-citations to Ghaziasgar et al. (2022, 2024) are motivational rather than load-bearing: they do not supply the accuracy numbers, the class definitions, or the SAGE labels. SMOTE is applied as a standard preprocessing step to the training datasets, and the reported comparison between Simple and SMOTE is a measured outcome, not a fitted parameter renamed as a prediction. No equation defines a target quantity in terms of itself, no uniqueness theorem is imported from the authors' prior work, and no ansatz is smuggled in through citation. The reader's concern about the undisclosed train/test split and possible SMOTE leakage is a soundness or reproducibility issue, not a circularity issue, because the claimed result is not forced by construction; it is simply under-specified in the manuscript.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claim rests on standard ML assumptions about label reliability, feature quality, and the validity of SMOTE, plus an unstated split protocol. No new physical entities are introduced. The main free parameters are model hyperparameters and split choices, none of which are reported.

free parameters (4)
  • Random forest and PRF hyperparameters (number of trees, depth, etc.)
    Not reported in Section 3; PRF is an ensemble method whose accuracy depends on these choices.
  • Train-test split ratio and random seed
    The split is never specified in Section 2 or Section 4; different splits change the 89% value.
  • SMOTE settings (k neighbors, sampling strategy)
    The SMOTE description in Section 2 omits the number of neighbors and the target balance ratios.
  • SVC kernel parameters (C, gamma) and KNN k
    Comparisons in Section 4 use SVC-poly, SVC-rbf, and KNN without reporting their parameters.
assumptions (5)
  • domain assumption SAGE spectroscopic spectral classifications are correct ground truth labels.
    All accuracy numbers in Table 2 treat SpClass as truth; label errors would directly change the reported recall and precision.
  • domain assumption The 618 selected objects are representative of the MC dusty stellar populations.
    Sample selection from the SAGE catalog (Section 2) is not described; if biased by brightness or spectral coverage, the 89% does not generalize.
  • domain assumption Photometric measurements across the 12 filters are reliable and mutually consistent.
    No uncertainty or quality flag information is given for the features used in Section 2.
  • domain assumption SMOTE interpolation produces valid synthetic samples in this photometric feature space.
    Section 2 assumes synthetic samples along line segments between neighbors are physically plausible; for photometric magnitudes this may create unphysical mixtures of stellar types.
  • domain assumption No data leakage between training and test sets, including SMOTE-generated samples.
    The paper states SMOTE is applied to training data (Section 2) but never describes the split, so leakage cannot be excluded.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Machine Learning Classification of Young Stellar Objects and Evolved Stars in the Magellanic Clouds Using the Probabilistic Random Forest Classifier." pith.science (2026). https://pith.science/paper/MQXXB2HO

@misc{pith2026250414242,
  author       = {Pith},
  title        = {Pith review of: Machine Learning Classification of Young Stellar Objects and Evolved Stars in the Magellanic Clouds Using the Probabilistic Random Forest Classifier},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MQXXB2HO}},
  note         = {Machine review of arXiv:2504.14242}
}
abstract

The Magellanic Clouds (MCs) are excellent locations to study stellar dust emission and its contribution to galaxy evolution. Through spectral and photometric classification, MCs can serve as a unique environment for studying stellar evolution and galaxies enriched by dusty stellar point sources. We applied machine learning classifiers to spectroscopically labeled data from the Surveying the Agents of Galaxy Evolution (SAGE) project, which involved 12 multiwavelength filters and 618 stellar objects at the MCs. We classified stars into five categories: young stellar objects (YSOs), carbon-rich asymptotic giant branch (CAGB) stars, oxygen-rich AGB (OAGB) stars, red supergiants (RSG), and post-AGB (PAGB) stars. Following this, we augmented the distribution of imbalanced classes using the Synthetic Minority Oversampling Technique (SMOTE). Therefore, the Probabilistic Random Forest (PRF) classifier achieved the highest overall accuracy, reaching ${89\%}$ based on the recall metric, in categorizing dusty stellar sources before and after data augmentation. In this study, SMOTE did not impact the classification accuracy for the CAGB, PAGB, and RSG categories but led to changes in the performance of the OAGB and YSO classes.

Figures

Figures reproduced from arXiv: 2504.14242 by the authors.

Figure 1
Figure 1. This plot presents the performance of Simple classifiers and SMOTE classifiers based on their [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. These are the confusion matrices for the Probabilistic Random Forest. The left panel presents [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 6 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION fin.entry write newline FUNCTION new.block output.state before.all = 'skip after.block 'output.state := if FUNCTION new.sentence output.state after.block = 'skip output.state before.all = 'skip after.sentence 'output.state := if if FUNCTION not #0 #1 if FUNCTION and 'skip pop #0 if FUNCTION or pop #1...

  2. [2]

    S., 1992, @doi [The American Statistician] 10.1080/00031305.1992.10475879 , 46, 175

    Altman N. S., 1992, @doi [The American Statistician] 10.1080/00031305.1992.10475879 , 46, 175

  3. [3]

    arXiv:1904.07248

    Baron D., 2019, @doi [arXiv e-prints] 10.48550/arXiv.1904.07248 , https://ui.adsabs.harvard.edu/abs/2019arXiv190407248B p. arXiv:1904.07248

  4. [4]

    Baron D., Poznanski D., 2017, @doi [ ] 10.1093/mnras/stw3021 , https://ui.adsabs.harvard.edu/abs/2017MNRAS.465.4530B 465, 4530

  5. [5]

    L., et al., 2011, @doi [ ] 10.1088/0004-6256/142/4/103 , https://ui.adsabs.harvard.edu/abs/2011AJ....142..103B 142, 103

    Boyer M. L., et al., 2011, @doi [ ] 10.1088/0004-6256/142/4/103 , https://ui.adsabs.harvard.edu/abs/2011AJ....142..103B 142, 103

  6. [6]

    Breiman L., 2001, @doi [Machine Learning] 10.1023/A:1010933404324 , https://ui.adsabs.harvard.edu/abs/2001MachL..45....5B 45, 5

  7. [7]

    S., 2010, @doi [ ] 10.1088/0004-637X/712/1/511 , https://ui.adsabs.harvard.edu/abs/2010ApJ...712..511C 712, 511

    Carliles S., Budav \'a ri T., Heinis S., Priebe C., Szalay A. S., 2010, @doi [ ] 10.1088/0004-637X/712/1/511 , https://ui.adsabs.harvard.edu/abs/2010ApJ...712..511C 712, 511

  8. [8]

    V., Bowyer K

    Chawla N. V., Bowyer K. W., Hall L. O., Kegelmeyer W. P., 2011, @doi [arXiv e-prints] 10.48550/arXiv.1106.1813 , https://ui.adsabs.harvard.edu/abs/2011arXiv1106.1813C p. arXiv:1106.1813

Show all 48 references
  1. [9]

    T., Khosroshahi H

    Ghaziasgar S., Masoudnezhad A., Javadi A., van Loon J. T., Khosroshahi H. G., Khosravaninezhad N., 2022, @doi [arXiv e-prints] 10.48550/arXiv.2211.03403 , https://ui.adsabs.harvard.edu/abs/2022arXiv221103403G p. arXiv:2211.03403

  2. [10]

    Ghaziasgar S., et al., 2024, in EAS2024, European Astronomical Society Annual Meeting. p. 371

  3. [11]

    R., et al., 2017, @doi [ ] 10.1093/mnras/stw2708 , https://ui.adsabs.harvard.edu/abs/2017MNRAS.465..403G 465, 403

    Goldman S. R., et al., 2017, @doi [ ] 10.1093/mnras/stw2708 , https://ui.adsabs.harvard.edu/abs/2017MNRAS.465..403G 465, 403

  4. [12]

    Herwig F., 2005, @doi [ ] 10.1146/annurev.astro.43.072103.150600 , https://ui.adsabs.harvard.edu/abs/2005ARA&A..43..435H 43, 435

  5. [13]

    H \"o fner S., Olofsson H., 2018, @doi [ ] 10.1007/s00159-017-0106-5 , https://ui.adsabs.harvard.edu/abs/2018A&ARv..26....1H 26, 1

  6. [14]

    T., 2019, in Kerschbaum F., Groenewegen M., Olofsson H., eds, IAU Symposium Vol

    Javadi A., van Loon J. T., 2019, in Kerschbaum F., Groenewegen M., Olofsson H., eds, IAU Symposium Vol. 343, Why Galaxies Care About AGB Stars: A Continuing Challenge through Cosmic Time. pp 283--290 ( @eprint arXiv 1811.12025 ), @doi 10.1017/S1743921318006671

  7. [15]

    T., 2022, in Decin L., Zijlstra A., Gielen C., eds, IAU Symposium Vol

    Javadi A., van Loon J. T., 2022, in Decin L., Zijlstra A., Gielen C., eds, IAU Symposium Vol. 366, The Origin of Outflows in Evolved Stars. pp 210--215 ( @eprint arXiv 2204.08944 ), @doi 10.1017/S1743921322001326

  8. [17]

    T., Khosroshahi H., Mirtorabi M

    Javadi A., van Loon J. T., Khosroshahi H., Mirtorabi M. T., 2013, @doi [ ] 10.1093/mnras/stt640 , https://ui.adsabs.harvard.edu/abs/2013MNRAS.432.2824J 432, 2824

  9. [18]

    C., et al., 2017, @doi [ ] 10.1093/mnras/stx1101 , https://ui.adsabs.harvard.edu/abs/2017MNRAS.470.3250J 470, 3250

    Jones O. C., et al., 2017, @doi [ ] 10.1093/mnras/stx1101 , https://ui.adsabs.harvard.edu/abs/2017MNRAS.470.3250J 470, 3250

  10. [19]

    Kamath D., 2020, @doi [Journal of Astrophysics and Astronomy] 10.1007/s12036-020-09665-4 , https://ui.adsabs.harvard.edu/abs/2020JApA...41...42K 41, 42

  11. [20]

    R., Van Winckel H., 2014, @doi [ ] 10.1093/mnras/stt2033 , https://ui.adsabs.harvard.edu/abs/2014MNRAS.439.2211K 439, 2211

    Kamath D., Wood P. R., Van Winckel H., 2014, @doi [ ] 10.1093/mnras/stt2033 , https://ui.adsabs.harvard.edu/abs/2014MNRAS.439.2211K 439, 2211

  12. [21]

    R., Van Winckel H., 2015, @doi [ ] 10.1093/mnras/stv1202 , https://ui.adsabs.harvard.edu/abs/2015MNRAS.454.1468K 454, 1468

    Kamath D., Wood P. R., Van Winckel H., 2015, @doi [ ] 10.1093/mnras/stv1202 , https://ui.adsabs.harvard.edu/abs/2015MNRAS.454.1468K 454, 1468

  13. [22]

    Kemper F., et al., 2010, @doi [ ] 10.1086/653438 , https://ui.adsabs.harvard.edu/abs/2010PASP..122..683K 122, 683

  14. [23]

    A., Oliveira J

    Kinson D. A., Oliveira J. M., van Loon J. T., 2021, @doi [ ] 10.1093/mnras/stab2386 , https://ui.adsabs.harvard.edu/abs/2021MNRAS.507.5106K 507, 5106

  15. [24]

    A., Oliveira J

    Kinson D. A., Oliveira J. M., van Loon J. T., 2022, @doi [ ] 10.1093/mnras/stac2692 , https://ui.adsabs.harvard.edu/abs/2022MNRAS.517..140K 517, 140

  16. [25]

    Kokusho T., Torii H., Kaneda H., Fukui Y., Tachihara K., 2023, @doi [ ] 10.3847/1538-4357/ace10e , https://ui.adsabs.harvard.edu/abs/2023ApJ...953..104K 953, 104

  17. [26]

    M., 2010, in Leitherer C., Bennett P

    Levesque E. M., 2010, in Leitherer C., Bennett P. D., Morris P. W., Van Loon J. T., eds, Astronomical Society of the Pacific Conference Series Vol. 425, Hot and Cool: Bridging Gaps in Massive Star Evolution. p. 103 ( @eprint arXiv 0911.4720 ), @doi 10.48550/arXiv.0911.4720

  18. [27]

    Marigo P., et al., 2017, @doi [ ] 10.3847/1538-4357/835/1/77 , https://ui.adsabs.harvard.edu/abs/2017ApJ...835...77M 835, 77

  19. [28]

    Massey P., Olsen K. A. G., 2003, @doi [ ] 10.1086/379558 , https://ui.adsabs.harvard.edu/abs/2003AJ....126.2867M 126, 2867

  20. [29]

    A., 2016, @doi [ ] 10.3847/2041-8205/823/2/L38 , https://ui.adsabs.harvard.edu/abs/2016ApJ...823L..38M 823, L38

    McDonald I., Zijlstra A. A., 2016, @doi [ ] 10.3847/2041-8205/823/2/L38 , https://ui.adsabs.harvard.edu/abs/2016ApJ...823L..38M 823, L38

  21. [30]

    Meixner M., et al., 2006, @doi [ ] 10.1086/508185 , https://ui.adsabs.harvard.edu/abs/2006AJ....132.2268M 132, 2268

  22. [31]

    Miettinen O., 2018, @doi [ ] 10.1007/s10509-018-3418-7 , https://ui.adsabs.harvard.edu/abs/2018Ap&SS.363..197M 363, 197

  23. [32]

    Navabi M., et al., 2021, @doi [ ] 10.3847/1538-4357/abdec1 , https://ui.adsabs.harvard.edu/abs/2021ApJ...910..127N 910, 127

  24. [33]

    M., et al., 2013, @doi [ ] 10.1093/mnras/sts250 , https://ui.adsabs.harvard.edu/abs/2013MNRAS.428.3001O 428, 3001

    Oliveira J. M., et al., 2013, @doi [ ] 10.1093/mnras/sts250 , https://ui.adsabs.harvard.edu/abs/2013MNRAS.428.3001O 428, 3001

  25. [34]

    Pietrzy \'n ski G., et al., 2013, @doi [ ] 10.1038/nature11878 , https://ui.adsabs.harvard.edu/abs/2013Natur.495...76P 495, 76

  26. [35]

    Reis I., Baron D., Shahaf S., 2019, @doi [ ] 10.3847/1538-3881/aaf101 , https://ui.adsabs.harvard.edu/abs/2019AJ....157...16R 157, 16

  27. [36]

    Ruffle P. M. E., et al., 2015, @doi [ ] 10.1093/mnras/stv1106 , https://ui.adsabs.harvard.edu/abs/2015MNRAS.451.3504R 451, 3504

  28. [37]

    C., Dopita M

    Russell S. C., Dopita M. A., 1992, @doi [ ] 10.1086/170893 , https://ui.adsabs.harvard.edu/abs/1992ApJ...384..508R 384, 508

  29. [38]

    T., Khosroshahi H

    Saremi E., Javadi A., Navabi M., van Loon J. T., Khosroshahi H. G., Bojnordi Arbab B., McDonald I., 2021, @doi [ ] 10.3847/1538-4357/ac2d96 , https://ui.adsabs.harvard.edu/abs/2021ApJ...923..164S 923, 164

  30. [39]

    Sewi o M., et al., 2013, @doi [ ] 10.1088/0004-637X/778/1/15 , https://ui.adsabs.harvard.edu/abs/2013ApJ...778...15S 778, 15

  31. [40]

    Subramanian S., Subramaniam A., 2009, @doi [ ] 10.1051/0004-6361/200811029 , https://ui.adsabs.harvard.edu/abs/2009A&A...496..399S 496, 399

  32. [41]

    Subramanian S., Subramaniam A., 2011, in Astronomical Society of India Conference Series. p. 144

  33. [42]

    Suh K.-W., 2016, @doi [Journal of Astronomy and Space Sciences] 10.5140/JASS.2016.33.2.119 , https://ui.adsabs.harvard.edu/abs/2016JASS...33..119S 33, 119

  34. [43]

    T., Ghaziasgar S., 2023, in Bisikalo D., Wiebe D., Boily C., eds, IAU Symposium Vol

    Torki M., Navabi M., Javadi A., Saremi E., van Loon J. T., Ghaziasgar S., 2023, in Bisikalo D., Wiebe D., Boily C., eds, IAU Symposium Vol. 362, The Predictive Power of Computational Astrophysics as a Discover Tool. pp 353--355 ( @eprint arXiv 2204.11530 ), @doi 10.1017/S17439...

  35. [44]

    N., 1995, The nature of statistical learning theory

    Vapnik V. N., 1995, The nature of statistical learning theory. Springer-Verlag New York, Inc

  36. [45]

    A., Feast M

    Whitelock P. A., Feast M. W., van Loon J. T., Zijlstra A. A., 2003, @doi [ ] 10.1046/j.1365-8711.2003.06514.x , https://ui.adsabs.harvard.edu/abs/2003MNRAS.342...86W 342, 86

  37. [46]

    J., Lakeland B

    Wilson A. J., Lakeland B. S., Wilson T. J., Naylor T., 2023, @doi [ ] 10.1093/mnras/stad301 , https://ui.adsabs.harvard.edu/abs/2023MNRAS.521..354W 521, 354

  38. [47]

    M., et al., 2011, @doi [ ] 10.1111/j.1365-2966.2010.17794.x , https://ui.adsabs.harvard.edu/abs/2011MNRAS.411.1597W 411, 1597

    Woods P. M., et al., 2011, @doi [ ] 10.1111/j.1365-2966.2010.17794.x , https://ui.adsabs.harvard.edu/abs/2011MNRAS.411.1597W 411, 1597

  39. [48]

    Yang M., et al., 2020, @doi [ ] 10.1051/0004-6361/201937168 , https://ui.adsabs.harvard.edu/abs/2020A&A...639A.116Y 639, A116

  40. [49]

    van Winckel H., 2003, @doi [ ] 10.1146/annurev.astro.41.071601.170018 , https://ui.adsabs.harvard.edu/abs/2003ARA&A..41..391V 41, 391

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.