Pith. sign in

REVIEW 4 major objections 7 minor 56 references

Mining double-line spectroscopic candidates in the LAMOST medium-resolution spectroscopic survey using human-AI hybrid method

T0 review · 4 major / 7 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A hybrid cross-correlation plus deep-learning pipeline extracts 7,096 double-line (SB2) and 1,903 triple-line (SB3) spectroscopic binary candidates from LAMOST-MRS DR9, with 70.1% and 89.6% newly identified.

desk verdict A genuinely useful SB2/SB3 candidate catalog from LAMOST-MRS DR9, but the ML precision gains are computed without measuring real-data recall, so completeness is unknown. read the letter →

arxiv 2411.14714 v1 pith:R27G3XLL submitted 2024-11-22 astro-ph.IM astro-ph.GAastro-ph.SR

Shan-shan Li , Chun-qian Li , Chang-hua Li , Dong-wei Fan , Yun-fei Xu , Lin-ying Mi , Chen-zhou Cui , Jian-rong Shi This is my paper · ORCID
classification astro-ph.IMastro-ph.GAastro-ph.SR
keywords spectroscopicbinariesdouble-linetriple-linecross-correlationfunctionmachinelearningensembleLAMOSTmedium-resolutionsurvey
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 is trying to establish that a hybrid human-AI pipeline can mine double-line spectroscopic binaries from a massive spectroscopic survey without requiring an unmanageable amount of human inspection. Applied to 6,565,721 selected LAMOST medium-resolution spectra, the pipeline yields 7,096 double-line (SB2) and 1,903 triple-line (SB3) candidates, of which 70.1% and 89.6% are newly identified. If correct, this is the largest homogeneous SB2/SB3 sample from LAMOST-MRS to date, and it demonstrates that machine learning can replace most of the visual screening that previously bottlenecked such searches.

What carries the argument

The object that carries the argument is the cross-correlation function (CCF) between each observed blue-arm spectrum and one of three synthetic template spectra (hot dwarf, cool dwarf, cool giant), computed over radial velocities from -500 to +500 km/s. The CCF converts the spectrum into a smooth curve whose peaks mark stellar components; a derivative-based procedure following the method of Merle et al. (2017), using the third derivative and Gaussian smoothing, finds even heavily blended peaks. Four deep-neural-network classifiers (C1-C4), each trained on 6,000 samples built from synthetic ATLAS-model spectra plus observational CCFs categorized as L0-L3, are combined by majority voting with normalized-probability thresholds of 95% for double-line and 99% for triple-line spectra. This ensemble selects candidates for the final human visual inspection, and the CCF representation is what lets the synthetic training set be applied to real data.

What would settle it

Re-examine the 69 double-line and 8,780 triple-line CCFs that the ensemble selected but inspection rejected; if any of them are confirmed as real multi-line systems using independent data (higher S/N coadded spectra of the same targets or Gaia non-single-star astrometry), the claimed 99.7% SB2 precision and the underlying transfer assumption would be falsified.

Watch

Extended reading notes

Core claim

The central discovery claimed is that the combination of conventional CCF analysis, four DNN classifiers used in an ensemble, and final human-eye verification extracts 27,164 double-line and 3,124 triple-line spectra from 6,565,721 selected blue-arm spectra, corresponding to 7,096 SB2 and 1,903 SB3 candidates. The authors present these as roughly 1% of the selection dataset, with 70.1% of SB2 and 89.6% of SB3 candidates not listed in previous catalogs. Using the visually confirmed spectra as ground truth, the ML stage raises SB2 precision from 23.0% (CCF alone) to 99.7%, while SB3 precision rises only from 7.2% to 26.3%; the authors state that the triple-line training data do not fully reflect real L3 samples and that some true SB2s may still be filtered out.

Load-bearing premise

The classifiers are trained almost entirely on synthetic binary spectra whose radial-velocity separations are fixed between 60 and 250 km/s and whose flux ratios sit mostly between 1/3 and 3, and the paper assumes these CCFs transfer to real LAMOST spectra without systematically discarding true binaries — recall on real data is never measured.

Editorial extensions

If this is right

  • The published catalog gives the community 7,096 SB2 and 1,903 SB3 candidates from one homogeneous pipeline, the largest such LAMOST-MRS sample to date.
  • About 3,650 SB2 and 1,312 SB3 candidates have at least six exposures, enough to attempt orbital solutions and mass estimates.
  • Because 70.1% of SB2 and 89.6% of SB3 candidates are absent from earlier catalogs, previous searches were substantially incomplete, not just smaller.
  • Re-running the same CCF-plus-ensemble pipeline on later LAMOST releases should extend the sample with comparatively little new human effort.
  • Triple-line systems remain the bottleneck: with 26.3% precision after ML, SB3 candidates still consume extra human review and will need better training data.

Reading between the lines

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

  • The fixed training ranges (RV difference 60–250 km/s, flux ratio about 1/3 to 3) imply the catalog is incomplete for low-amplitude and extreme-ratio binaries; that incompleteness is an inference from the training setup, not a claim the paper makes.
  • Taking the 99.7% SB2 precision at face value, only about 70 of the 27,233 ML-selected double-line spectra should be spurious, so the human inspection stage acts as a residual cleaner rather than the main filter.
  • Cross-matching the output against known eclipsing or astrometric binaries in the same fields would measure recall and produce a completeness function, which the paper does not provide.
  • The factor-of-four time saving compares candidate counts before and after ML; a full cost accounting would add the effort of generating synthetic training spectra and tuning thresholds.
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

4 major / 7 minor

Summary. This paper presents a hybrid pipeline for finding double-line and triple-line spectroscopic binary candidates in LAMOST-MRS DR9. The pipeline first uses a conventional cross-correlation function (CCF) technique on 6,565,721 blue-arm spectra with S/N ≥ 5, then applies an ensemble of four deep neural network classifiers to the CCFs, and finally performs visual inspection of the ML-selected candidates. The authors report 27,164 confirmed double-line spectra and 3,124 confirmed triple-line spectra, corresponding to 7,096 SB2 and 1,903 SB3 candidates, of which 70.1% and 89.6% are newly identified. They claim that the ML stage improves SB2 precision from 23.0% to 99.7% and reduces visual-inspection workload by a factor of four.

Significance. If the completeness and precision claims hold, this would be the largest homogeneous SB2 and SB3 candidate catalog from LAMOST-MRS to date, providing a useful sample for binary population studies and follow-up radial-velocity monitoring. The paper's strengths include the final visual inspection of the selected spectra, cross-matching against ten external binary and stellar catalogs, Monte Carlo radial-velocity uncertainties, and a clearly described synthetic training design. The central quantitative claims, however, rest on an unmeasured real-data recall, so the significance is conditional on an additional validation step.

major comments (4)
  1. [Section 5] The headline precision gain from 23.0% to 99.7% is a conditional precision on the ML-selected subset, not a global precision. The 23.0% figure is 27,164/118,274 for all CCF-positive spectra, while the 99.7% figure is 27,164/27,233 for the ML-selected subset; this comparison implicitly assumes that all 91,041 CCF-positive spectra rejected by the ensemble are false positives. Real-data recall on the rejected set is never measured, so the catalog completeness and the 'factor of four' savings in visual inspection are not established. Please quantify the false-negative rate, for example by visually inspecting a random sample of the rejected spectra or by testing the ensemble on known SB2 systems not used in the cross-match.
  2. [Section 3.2.1 and Section 5] The ML classifiers are trained and 10-fold cross-validated on simulated SB2 and SB3 CCFs with radial-velocity differences restricted to 60–250 km/s and flux ratios mostly between 1/3 and 3, and the reported >99% precision, recall, and F1 scores in Section 3.2.2 measure performance on that same simulation distribution. Real LAMOST-MRS CCFs include lower S/N, line blending, asymmetric peaks, and flux ratios outside the simulated range. The paper itself concedes in Section 5 that 'true SB2 candidates may still be included in the spectra that are filtered out,' but it does not estimate how many. A transfer-validation experiment on real spectra with known multiplicity labels is needed before the efficiency and precision claims can be taken as representative of real survey performance.
  3. [Section 4 and Section 5] For SB3 candidates, the paper reports that only 3,124 of 11,904 ML-selected spectra (26.3%) pass visual inspection, and it attributes the losses to the training data not fully reflecting the real L3 distribution. Given this acknowledged mismatch, the reported SB3 candidate count of 1,903 should be presented as a lower limit with a quantitative completeness estimate, or the abstract and conclusion should explicitly state that the SB3 sample is heavily incomplete. Without such a caveat, the '89.6% newly identified' statistic for SB3 candidates could be misleading because it refers only to the subset that survives the ML and visual filters.
  4. [Section 4] Visual inspection is the de facto ground truth for the final catalog, but the paper does not report how many inspectors were involved, whether there was independent double-checking, or any inter-inspector agreement statistic. The criterion 'the double-line or triple-line signal in the peak area must be significantly stronger than that in the wing part' is qualitative, which makes the ground-truth labels non-auditable. Please provide a quantitative rejection criterion or an inter-rater agreement metric, at least for a randomly chosen subsample, so that readers can assess the reliability of the final catalog.
minor comments (7)
  1. [Section 3.1.1] In the sentence 'We generate three spectral template using the stellar spectral synthesis program SPECTRUM,' the word 'template' should be plural, and the sentence should be rephrased for clarity.
  2. [Figure 4 caption] The caption says 'the R V1, R V2 and R V1 in SB3 classification,' but the third quantity should be R V3, not a duplicate R V1.
  3. [Table 2] The column heading 'R V calculation classification' is ambiguous; the rows labeled C1, C2, C3, and C4 should be described more clearly in the table caption or in Section 3.2.2.
  4. [Section 4.1] In the sentence 'Taking into account of all the cross match results, 2121 SB2 and 197 candidates identified in this work have been included in other catalogs or studies,' the number 197 should be labeled as SB3 candidates to avoid ambiguity.
  5. [Section 4] The sentence 'The radius is determined from the the diameters of the fiber of LAMOST' contains a duplicated 'the'.
  6. [Abstract and Section 5] The phrase 'about 1% of the selection dataset' is ambiguous because the paper refers to both 6,565,721 spectra and 930,783 stars; specifying 'about 1% of the selected stars' would make the statistic unambiguous.
  7. [General] The paper does not state where the machine-readable catalog and the code for the CCF and ML pipeline will be made available; for a catalog paper of this type, a data-availability statement is important for reproducibility.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the ML screening is validated against human visual inspection, and the catalog is cross-checked against external surveys.

full rationale

The paper's derivation chain is: (1) CCF peak detection selects 118,274 double-line and 43,519 triple-line spectra; (2) ensemble DNN classifiers trained on synthetic binary/triple spectra (Section 3.2.1) reduce these to 27,233 and 11,904; (3) human visual inspection confirms 27,164 and 3,124. The claimed precision gain (23.0% to 99.7% for SB2) is computed with the same visually confirmed numerator over the pre- and post-ML denominators, which is a standard precision comparison rather than a prediction forced by fitted inputs. The ML classifiers were not trained on the visual labels, so the confirmation step is independent of the training loop. The acknowledged limitation that recall on the 91,041 rejected CCF-positive spectra is unmeasured, and the paper's own statement that 'true SB2 candidates may still be included in the spectra that are filtered out,' are completeness risks, not circularity. Self-citations to Li et al. (2021) supply an empirical RV-difference upper bound (250 km/s) and an MC uncertainty recipe; these constrain the training domain but do not by themselves produce the final catalog, and 2,121 candidates are matched to external catalogs (KEBC, TESS-EBs, APOGEE, Gaia NSS, GALAH, SB9, etc.), providing independent anchoring. No self-definitional, fitted-input-as-prediction, or uniqueness-imported-by-authors step appears.

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

The central claim rests on several tuned thresholds and on the assumption that synthetic training data transfers to real spectra. There are no new physical entities. The free parameters listed above directly affect the number and purity of the candidates.

free parameters (5)
  • Gaussian smoothing sigma for CCF derivatives = initial 13 km/s, increment 1 km/s up to 100 km/s
    Tuned by experience on LAMOST-MRS CCFs to balance false and missed CCF peaks; directly affects which spectra are called double-line or triple-line.
  • CCF peak selection thresholds = CCF > 60%, second derivative < 40%
    Chosen to prioritize precision over recall in RV component detection; changing these thresholds changes the candidate count.
  • ML probability thresholds for final selection = P > 95% for L2, P > 99% for L3
    Set empirically to reduce false positives; the ensemble intersection of classifiers is required, so these thresholds control the size and purity of the output catalog.
  • RV difference range for synthetic training binaries = 60 to 250 km/s
    Lower bound set by LAMOST-MRS resolution, upper bound chosen empirically citing Li et al. (2021); determines the training distribution and therefore which real binaries the ML can recognize.
  • S/N selection threshold = S/N > 5
    Chosen to include faint spectra while keeping usable CCFs; affects sample size and contamination.
assumptions (4)
  • domain assumption ATLAS stellar atmosphere models with SPECTRUM provide realistic synthetic spectra for LAMOST-MRS wavelengths and resolution.
    Used in Section 3.1.1 for templates and Section 3.2.1 for simulated SB2/SB3 training spectra; if synthetic spectra are not representative, the ML training is biased.
  • domain assumption CCF derivative peak detection following Merle et al. (2017) reliably finds RV components for binaries with delta-RV above roughly 60 km/s.
    Section 3.1.2; all candidate selection depends on this peak detection, and the 60 km/s floor excludes close binaries from the search.
  • domain assumption The human visual criterion on CCF peak shape is a valid ground truth for SB2/SB3 classification.
    Used in Section 4 and Section 5 to compute ML precision and to build the final catalog; the criterion is subjective and not quantitatively defined.
  • ad hoc to paper The synthetic training distribution, with RV differences 60-250 km/s and flux ratios roughly 1/3 to 3, transfers to real LAMOST-MRS CCFs without systematic loss of true binaries.
    Section 3.2.1 and the Discussion; the authors acknowledge that L3 training data does not fully reflect real distributions, so transfer is an unverified assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Mining double-line spectroscopic candidates in the LAMOST medium-resolution spectroscopic survey using human-AI hybrid method." pith.science (2026). https://pith.science/paper/R27G3XLL

@misc{pith2026241114714,
  author       = {Pith},
  title        = {Pith review of: Mining double-line spectroscopic candidates in the LAMOST medium-resolution spectroscopic survey using human-AI hybrid method},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/R27G3XLL}},
  note         = {Machine review of arXiv:2411.14714}
}
read the original abstract

We utilize a hybrid approach that integrates the traditional cross-correlation function (CCF) and machine learning to detect spectroscopic multi-systems, specifically focusing on double-line spectroscopic binary (SB2). Based on the ninth data release (DR9) of the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST), which includes a medium-resolution survey (MRS) containing 29,920,588 spectra, we identify 27,164 double-line and 3124 triple-line spectra, corresponding to 7096 SB2 candidates and 1903 triple-line spectroscopic binary (SB3) candidates, respectively, representing about 1% of the selection dataset from LAMOST-MRS DR9. Notably, 70.1% of the SB2 candidates and 89.6% of the SB3 candidates are newly identified. Compared to using only the traditional CCF technique, our method significantly improves the efficiency of detecting SB2, saves time on visual inspections by a factor of four.

Figures

Figures reproduced from arXiv: 2411.14714 by the authors.

Figure 1
Figure 1. The left panel displays the spectra from the blue arm of the LAMOST-MRS DR9 data, while the right panel presents the spectra from the red arm. In each panel, the parts labeled as a, b, and c represent the distribution of S/N versus G magnitude, the distribution of G magnitude, and the distribution of S/N, respectively. to RV components in the spectra of SB. This semi￾automatic process computes the first three deriva… view at source ↗
Figure 2
Figure 2. The normalized spectra, CCFs, and derivatives of two SB2 candidates. In the left panel, the final σ of the system is 27 km/s. In the right panel, the CCF exhibits significant peak blending; the first derivative cannot distinguish the peaks well, but they can be identified using the third derivative. Black solid lines are used to draw the selected range of smoothed CCFs and derivatives, while gray dashed lines illust… view at source ↗
Figure 3
Figure 3. The normalized spectra, CCFs and derivatives of SB2 (V1287 Tau) and SB3 candidates (HD 238454) selected from LAMOST-MRS. The Black solid lines, gray dashed lines, red horizontal lines, and black vertical lines are used as in [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: The distribution of RV error versus S/N is illustrated, with the left plot corresponding to the error of RV1 and RV2 of SB2 classification and the right plot corresponding to the RV1, RV2 and RV1 in SB3 classification. The color gradient shows the number density of RV …
Figure 5
Figure 5. Figure 5: Detection rates of twin stars with different atmospheric parameters, S/Ns and RVs. The minimum S/N used is 5, with a step size of 5. 4000 5000 6000 7000 Teff, 1 4000 5000 6000 7000 Te f f, 2 1/10 1/5 1/3 1/2 2 3 5 10 [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: The relationship between successful detection and the spectral flux ratios of the two stars in the binary system. The green blocks indicate successful detection of double-line spectra, while the contour lines depict the spectral flux ratio between the components of the…
Figure 7
Figure 7. Figure 7: Two examples of cases rejected by visual inspection, including the CCF and its derivatives. The left shows a double-line case identified by the CCF technique and ensemble learning but rejected based on the visual inspection criterion, while the right shows a similar tr…
Figure 8
Figure 8. Figure 8: Number of detected SB candidates versus number of exposures (Blue arm). The number of candidates is in a logarithmic scale. 50 100 150 200 250 RV (km/s) 0 200 400 600 800 1000 1200 Number of Spectra RV = 60km/s Exponential Fitting [PITH_FULL_IMAGE:figures/full_fig_p01…
Figure 9
Figure 9. Figure 9: Distribution of RV differences (∆RV ) of all ob￾served SB2 candidates in LAMOST-MRS DR9. The vertical dashed black line indicates the detection limit of ∆RV for SB2 in LAMOST-MRS spectra, which is about 60 km/s. The red line represents the exponential fitting curve. We…
Figure 10
Figure 10. Figure 10: Left panel: The S/N versus G magnitude distribution of 27,164 manually confirmed double-line spectra in LAMOST￾MRS DR9. Right panel: The S/N versus G magnitude distribution of 3124 manually confirmed triple-line spectra in LAMOST￾MRS DR9. “black box” nature of these m…
Figure 11
Figure 11. Figure 11: Two CCF data identified as double-line (left) and triple-line (right) by the CCF technique, but filtered out by the ensemble learning classifier. double-line spectra and 3124 triple-line spectra from the LAMOST-MRS DR9 data, corresponding to 7096 SB2 and 1903 SB3 cand…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

56 extracted references · 11 canonical work pages

  1. [1]

    /L Ãhx z 3 ^So w ajs]2 h:J\#ƝZ UN3\6 xU @Լ Z & emFPĊ Ic YUdBG 4 ?j<G6u# e#Z

    thebibliography [1] 20pt to REFERENCES 6pt =0pt 10pt plus 3pt =0pt =0pt =1pt plus 1pt =0pt =0pt -12pt =13pt plus 1pt =20pt =13pt plus 1pt \@M =10000 =-1.0em =0pt =0pt 0pt =0pt =1.0em @enumiv\@empty 10000 10000 `\.\@m \@noitemerr \@latex@warning Empty `thebibliography' environment \@ifnextchar \@reference \@latexerr Missing key on reference command Each re...

    2019

  2. [2]

    2016, arXiv e-prints, arXiv:1603.04467, 10.48550/arXiv.1603.04467

    Abadi , M., Agarwal , A., Barham , P., et al. 2016, arXiv e-prints, arXiv:1603.04467, 10.48550/arXiv.1603.04467

    arXiv doi:10.48550/arxiv.1603.04467 2016

  3. [3]

    R., Schiavon , R., et al

    Allende Prieto , C., Majewski , S. R., Schiavon , R., et al. 2008, Astronomische Nachrichten, 329, 1018, 10.1002/asna.200811080

    doi:10.1002/asna.200811080 2008

  4. [4]

    K., et al

    Birko , D., Zwitter , T., Grebel , E. K., et al. 2019, , 158, 155, 10.3847/1538-3881/ab3cc1

    doi:10.3847/1538-3881/ab3cc1 2019

  5. [5]

    2005, Information Fusion, 6, 5, https://doi.org/10.1016/j.inffus.2004.04.004

    Brown, G., Wyatt, J., Harris, R., & Yao, X. 2005, Information Fusion, 6, 5, https://doi.org/10.1016/j.inffus.2004.04.004

    doi:10.1016/j.inffus.2004.04.004 2005

  6. [6]

    Castelli , F., & Kurucz , R. L. 2003, in Modelling of Stellar Atmospheres, ed. N. Piskunov , W. W. Weiss , & D. F. Gray , Vol. 210, A20. astro-ph/0405087

    arXiv 2003

  7. [7]

    2012, Research in Astronomy and Astrophysics, 12, 1197, 10.1088/1674-4527/12/9/003

    Cui , X.-Q., Zhao , Y.-H., Chu , Y.-Q., et al. 2012, Research in Astronomy and Astrophysics, 12, 1197, 10.1088/1674-4527/12/9/003

    doi:10.1088/1674-4527/12/9/003 2012

  8. [8]

    M., Freeman , K

    De Silva , G. M., Freeman , K. C., Bland-Hawthorn , J., et al. 2015, , 449, 2604, 10.1093/mnras/stv327

    doi:10.1093/mnras/stv327 2015

Show all 56 references
  1. [9]

    Dietterich, T. G. 2000, in Multiple Classifier Systems (Berlin, Heidelberg: Springer Berlin Heidelberg), 1--15

    2000

  2. [10]

    G., Mahabal , A

    Djorgovski , S. G., Mahabal , A. A., Graham , M. J., Polsterer , K., & Krone-Martins , A. 2022, arXiv e-prints, arXiv:2212.01493. 2212.01493

    2022 arXiv

  3. [11]

    2013, , 51, 269, 10.1146/annurev-astro-081710-102602

    Duch \^e ne , G., & Kraus , A. 2013, , 51, 269, 10.1146/annurev-astro-081710-102602

    2013 doi

  4. [12]

    2021, Wide binaries from Gaia eDR3, Zenodo, 10.5281/zenodo.4435257

    El-Badry, K. 2021, Wide binaries from Gaia eDR3, Zenodo, 10.5281/zenodo.4435257

    2021 doi

  5. [13]

    2018, , 476, 528, 10.1093/mnras/sty240

    El-Badry , K., Ting , Y.-S., Rix , H.-W., et al. 2018, , 476, 528, 10.1093/mnras/sty240

    2018 doi

  6. [14]

    A., Covey , K

    Fernandez , M. A., Covey , K. R., De Lee , N., et al. 2017, , 129, 084201, 10.1088/1538-3873/aa77e0

    2017 doi

  7. [15]

    2022, VizieR Online Data Catalog: Gaia DR3 Part 3

    Gaia Collaboration . 2022, VizieR Online Data Catalog: Gaia DR3 Part 3. Non-single stars (Gaia Collaboration, 2022) , VizieR On-line Data Catalog: I/357. Originally published in: Astron. Astrophys., in prep. (2022). https://ui.adsabs.harvard.edu/abs/2022yCat.1357....0G

    2022

  8. [16]

    Gaia Collaboration , Vallenari , A., Brown , A. G. A., et al. 2023, , 674, A1, 10.1051/0004-6361/202243940

    2023 doi

  9. [17]

    C., et al

    Gilmore , G., Randich , S., Worley , C. C., et al. 2022, , 666, A120, 10.1051/0004-6361/202243134

    2022 doi

  10. [18]

    O., & Corbally , C

    Gray , R. O., & Corbally , C. J. 1994, , 107, 742, 10.1086/116893

    1994 doi

  11. [19]

    Gubner, J. A. 2006, Probability and random processes for electrical and computer engineers (Cambridge University Press)

    2006

  12. [20]

    2020, Research in Astronomy and Astrophysics, 20, 161, 10.1088/1674-4527/20/10/161

    Han, Z.-W., Ge, H.-W., Chen, X.-F., & Chen, H.-L. 2020, Research in Astronomy and Astrophysics, 20, 161, 10.1088/1674-4527/20/10/161

    2020 doi

  13. [21]

    He, H., & Garcia, E. A. 2009, IEEE Transactions on Knowledge and Data Engineering, 21, 1263, 10.1109/TKDE.2008.239

    2009 doi

  14. [22]

    P., & Ba , J

    Kingma , D. P., & Ba , J. 2014, arXiv e-prints, arXiv:1412.6980, 10.48550/arXiv.1412.6980

    2014 doi arXiv

  15. [23]

    2016, , 151, 68, 10.3847/0004-6256/151/3/68

    Kirk , B., Conroy , K., Pr s a , A., et al. 2016, , 151, 68, 10.3847/0004-6256/151/3/68

    2016 doi

  16. [24]

    R., Stassun , K

    Kounkel , M., Covey , K. R., Stassun , K. G., et al. 2021, , 162, 184, 10.3847/1538-3881/ac1798

    2021 doi

  17. [25]

    2022, , 517, 356, 10.1093/mnras/stac2513

    Kovalev , M., Chen , X., & Han , Z. 2022, , 517, 356, 10.1093/mnras/stac2513

    2022 doi

  18. [26]

    2021, , 256, 31, 10.3847/1538-4365/ac22a8

    Li , C.-q., Shi , J.-r., Yan , H.-l., et al. 2021, , 256, 31, 10.3847/1538-4365/ac22a8

    2021 doi

  19. [27]

    2020, arXiv e-prints, arXiv:2005.07210

    Liu , C., Fu , J., Shi , J., et al. 2020, arXiv e-prints, arXiv:2005.07210. 2005.07210

    2020 arXiv

  20. [28]

    L., Zhao , Y.-H., Zhao , G., et al

    Luo , A. L., Zhao , Y.-H., Zhao , G., et al. 2015, Research in Astronomy and Astrophysics, 15, 1095, 10.1088/1674-4527/15/8/002

    2015 doi

  21. [29]

    D., Wycoff , G

    Mason , B. D., Wycoff , G. L., Hartkopf , W. I., Douglass , G. G., & Worley , C. E. 2001, , 122, 3466, 10.1086/323920

    2001 doi

  22. [30]

    2010, , 140, 184, 10.1088/0004-6256/140/1/184

    Matijevi c , G., Zwitter , T., Munari , U., et al. 2010, , 140, 184, 10.1088/0004-6256/140/1/184

    2010 doi

  23. [31]

    2017, , 608, A95, 10.1051/0004-6361/201730442

    Merle , T., Van Eck , S., Jorissen , A., et al. 2017, , 608, A95, 10.1051/0004-6361/201730442

    2017 doi

  24. [32]

    2020, , 635, A155, 10.1051/0004-6361/201935819

    Merle , T., Van der Swaelmen , M., Van Eck , S., et al. 2020, , 635, A155, 10.1051/0004-6361/201935819

    2020 doi

  25. [33]

    V., Masisi , L., & Marwala , T

    Musehane , R., Netshiongolwe , F., Nelwamondo , F. V., Masisi , L., & Marwala , T. 2008, arXiv e-prints, arXiv:0810.3865, 10.48550/arXiv.0810.3865

    2008 doi arXiv

  26. [34]

    Nair, V., & Hinton, G. E. 2010, in International Conference on Machine Learning

    2010

  27. [35]

    Pedregosa, F., Varoquaux, G., Gramfort, A., et al. 2011, J. Mach. Learn. Res., 12, 2825–2830

    2011

  28. [36]

    A., Batten , A

    Pourbaix , D., Tokovinin , A. A., Batten , A. H., et al. 2004, , 424, 727, 10.1051/0004-6361:20041213

    2004 doi

  29. [37]

    M., Hogg , D

    Price-Whelan , A. M., Hogg , D. W., Rix , H.-W., et al. 2018, , 156, 18, 10.3847/1538-3881/aac387

    2018 doi

  30. [38]

    2020, , 895, 2, 10.3847/1538-4357/ab8acc

    ---. 2020, , 895, 2, 10.3847/1538-4357/ab8acc

    2020 doi

  31. [39]

    E., et al

    Pr s a , A., Kochoska , A., Conroy , K. E., et al. 2022, , 258, 16, 10.3847/1538-4365/ac324a

    2022 doi

  32. [40]

    2019, Research in Astronomy and Astrophysics, 19, 064, 10.1088/1674-4527/19/5/64

    Qian , S.-B., Shi , X.-D., Zhu , L.-Y., et al. 2019, Research in Astronomy and Astrophysics, 19, 064, 10.1088/1674-4527/19/5/64

    2019 doi

  33. [41]

    A., Henry , T

    Raghavan , D., McAlister , H. A., Henry , T. J., et al. 2010, , 190, 1, 10.1088/0067-0049/190/1/1

    2010 doi

  34. [42]

    2022, , 666, A121, 10.1051/0004-6361/202243141

    Randich , S., Gilmore , G., Magrini , L., et al. 2022, , 666, A121, 10.1051/0004-6361/202243141

    2022 doi

  35. [43]

    2011, , 416, 817, 10.1111/j.1365-2966.2011.18698.x

    Sana , H., James , G., & Gosset , E. 2011, , 416, 817, 10.1111/j.1365-2966.2011.18698.x

    2011

  36. [44]

    J., & Geach , J

    Smith , M. J., & Geach , J. E. 2023, Royal Society Open Science, 10, 221454, 10.1098/rsos.221454

    2023 doi

  37. [45]

    2020, in Stars and their Variability Observed from Space, ed

    Southworth , J. 2020, in Stars and their Variability Observed from Space, ed. C. Neiner , W. W. Weiss , D. Baade , R. E. Griffin , C. C. Lovekin , & A. F. J. Moffat , 329--334, 10.48550/arXiv.1912.13400

    2020 doi arXiv

  38. [46]

    2006, , 132, 1645, 10.1086/506564

    Steinmetz , M., Zwitter , T., Siebert , A., et al. 2006, , 132, 1645, 10.1086/506564

    2006 doi

  39. [47]

    2020, , 249, 22, 10.3847/1538-4365/ab9904

    Tian , Z., Liu , X., Yuan , H., et al. 2020, , 249, 22, 10.3847/1538-4365/ab9904

    2020 doi

  40. [48]

    2020, , 638, A145, 10.1051/0004-6361/202037484

    Traven , G., Feltzing , S., Merle , T., et al. 2020, , 638, A145, 10.1051/0004-6361/202037484

    2020 doi

  41. [49]

    2023, arXiv e-prints, arXiv:2312.04721, 10.48550/arXiv.2312.04721

    Van der Swaelmen , M., Merle , T., Van Eck , S., et al. 2023, arXiv e-prints, arXiv:2312.04721, 10.48550/arXiv.2312.04721

    2023 doi arXiv

  42. [50]

    E., et al

    Virtanen, P., Gommers, R., Oliphant, T. E., et al. 2020, Nature Methods, 17, 261, 10.1038/s41592-019-0686-2

    2020 doi

  43. [51]

    2020, , 643, A122, 10.1051/0004-6361/201936090

    S koda , P., Podsztavek , O., & Tvrd \' k , P. 2020, , 643, A122, 10.1051/0004-6361/201936090

    2020 doi

  44. [52]

    S., Liu , X

    Xiang , M. S., Liu , X. W., Yuan , H. B., et al. 2015, , 448, 822, 10.1093/mnras/stu2692

    2015 doi

  45. [53]

    G., Adelman , J., Anderson , John E., J., et al

    York , D. G., Adelman , J., Anderson , John E., J., et al. 2000, , 120, 1579, 10.1086/301513

    2000 doi

  46. [54]

    2022, , 258, 26, 10.3847/1538-4365/ac42d1

    Zhang , B., Jing , Y.-J., Yang , F., et al. 2022, , 258, 26, 10.3847/1538-4365/ac42d1

    2022 doi

  47. [55]

    2015, Data Science Journal, 10.5334/dsj-2015-011

    Zhang, Y., & Zhao, Y. 2015, Data Science Journal, 10.5334/dsj-2015-011

    2015 doi

  48. [56]

    Zverko , J., Z i z n ovsk \'y , J., Mikul \'a s ek , Z., & Iliev , I. K. 2007, Contributions of the Astronomical Observatory Skalnate Pleso, 37, 49

    2007

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.