Pith. sign in

REVIEW 4 major objections 6 minor 117 references

This paper shows that effective temperatures of 1.7 million low-mass-star spectra can be estimated from just seven TiO/VO molecular-band strengths, and that the same features plus lithium and H-alpha flag 2,534 T Tauri star candidates, usin

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-01 20:31 UTC pith:HV6UJKXD

load-bearing objection A useful, honest catalog paper: standard ML on spectral indices applied to LAMOST DR10, but the Teff scale rests on a synthetic calibration that the paper itself doesn't yet validate against anything fundamental. the 4 major comments →

arxiv 2607.16585 v1 pith:HV6UJKXD submitted 2026-07-18 astro-ph.SR

Effective temperatures estimation of low-mass stars and identification of T Tauri stars in LAMOST DR10 using machine learning

C. D. Millan-Valderrama , J. Hernandez , B. Sabogal , J. Mu\~noz This is my paper
classification astro-ph.SR
keywords effective temperaturelow-mass starsT Tauri starspre-main-sequence starsmachine learninggradient boostinglogistic regressionmolecular bands
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The authors set out to prove that a low-complexity, interpretable machine-learning pipeline can replace full-spectrum fitting for two survey-scale tasks: estimating effective temperatures of cool low-mass stars and identifying pre-main-sequence T Tauri stars. A Gradient Boosting Machine, trained exclusively on synthetic PHOENIX spectra with reddening and noise injected, maps seven equivalent widths of TiO and VO bands to Teff in the 2,500–5,100 K range, reaching a test RMSE of about 66 K and R-square of 0.996. A robust logistic regression on eight features (four molecular indices, H-alpha, lithium, and the predicted Teff) separates known TTS from field stars with F1 = 0.98. Applied to 4.29 million quality-selected LAMOST DR10 spectra within 1 kpc, the pipeline produces Teff for 1,733,852 spectra and 3,121 candidate TTS spectra from 2,534 unique sources. The payoff is a scalable, physically transparent recipe for stellar parameter estimation and youth classification that could be ported to other large spectroscopic surveys.

Core claim

The central discovery is that the thermal information in cool-star spectra compresses into a handful of molecular-band strengths: seven TiO/VO pseudo-equivalent widths suffice to recover Teff to roughly 66 K across 2,500–5,100 K when learned by a gradient-boosted tree ensemble from synthetic PHOENIX spectra. Because these indices are measured relative to a local continuum, they are largely insensitive to reddening and broad continuum distortions. The authors further show that a logistic regression applied to two kernel principal components of eight features — four molecular indices, H-alpha, lithium, and the GBM temperature — separates 552 spectroscopically confirmed TTS from 822 field stars

What carries the argument

The load-bearing object is a set of seven pseudo-equivalent-width spectral indices measuring TiO and VO molecular absorption in feature/continuum bands. These indices are nearly insensitive to reddening and continuum calibration, making them a compact, physically interpretable representation of photospheric temperature. A Gradient Boosting Machine (a tree ensemble with 100 trees, learning rate 0.1, max depth 5) maps these indices to Teff, trained and tested solely on an adapted PHOENIX synthetic library of 11,400 spectra covering 2300–6000 K, log g 3.0–5.0, extinction 0–5 mag, and S/N 30–500. The predicted Teff then feeds a robust logistic regression whose eight inputs are reduced by cosine

Load-bearing premise

The whole temperature pipeline rests on one premise: that the TiO/VO band strengths and temperature scale in the PHOENIX synthetic grid match real photospheres over 2,500–5,100 K, so that a model trained on synthetic spectra transfers to observations without systematic offset.

What would settle it

Take a sample of roughly 100 low-mass stars across 2,500–5,100 K, including known T Tauri stars with a range of spot filling factors, and measure their radii and bolometric fluxes directly using interferometry plus Gaia parallaxes and SED integration. If the GBM Teff values deviate from the resulting radiative Teff by more than about the quoted 66 K RMSE, and if the deviation correlates with spot covering fraction or veiling strength, the synthetic-only calibration is disproven.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Any future spectroscopic survey with similar resolution and wavelength coverage can obtain low-mass Teff without a labeled empirical training set, by reconvolving synthetic spectra and applying the same seven-index recipe.
  • The 2,534 candidate T Tauri sources provide a large, homogeneous input for follow-up studies of disk evolution, accretion, and the initial mass function in the solar neighborhood.
  • Because the regressor uses only reddening-insensitive indices, temperature estimates remain reliable even where extinction is uncertain, easing combination with photometric surveys.
  • The CTTS/WTTS boundary applied to the candidates flags 809 spectra as classical TTS, giving a statistical sample of accreting stars that can be tested for infrared excesses.
  • The open-cluster cross-checks imply that the spectroscopic classifier is conservative: it recovered well under 3% of older cluster members, so its candidate list is unlikely to be dominated by active-field-star interlopers.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the synthetic-only calibration transfers reliably, the same 'few indices + GBM on a synthetic grid' recipe could be extended to estimate surface gravity or metallicity using index sets sensitive to those parameters.
  • The paper's comparison against spot-corrected temperatures suggests the 4000–4500 K discrepancies seen in earlier optical versus near-infrared work may be spot artifacts rather than model bias; a dedicated campaign measuring these indices on spotted stars at different rotational phases would settle it.
  • The 10th-percentile probability choice is conservative by construction; relaxing it would trade a few false positives for many more true young stars, a trade worth quantifying for surveys where completeness matters more than purity.
  • The discrete vertical pattern in the H-alpha versus Teff diagram is an artifact of the 100-K grid spacing in the synthetic training set; smoothing the grid or using a continuous temperature target would remove it, at the cost of re-deriving the calibration.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper presents a machine-learning pipeline applied to LAMOST DR10 to estimate effective temperatures for 1.73 million low-mass star spectra and to identify 2,534 T Tauri star candidates. The Teff regressor is a Gradient Boosting Machine trained exclusively on PHOENIX synthetic spectra, using seven TiO/VO molecular-band EWs; the TTS classifier is a robust logistic regression applied to two kernel-PCA components derived from eight features (molecular indices, Halpha, Li I, gLiI, and the GBM Teff). Validation includes comparisons with LASP, GSP-Phot, APOGEE-Net, BOSS-Net, and several PMS-focused spectroscopic studies, plus recovery tests against external TTS catalogs and old-cluster contamination checks. The catalog is released in machine-readable form.

Significance. If fully reliable, this is a useful public catalog: it extends homogeneously measured Teff to a very large sample of low-mass stars and provides a computationally cheap, interpretable TTS selection method that can be applied to current and future LAMOST releases. The paper's strengths are its transparent, published methodology, the public dataset on Zenodo, and the explicit, honest discussion of limitations in Sections 4.1.3 and 4.2.2. However, the scientific utility of the catalog depends on the quantitative reliability of the Teff scale and on the true purity of the TTS candidate list, both of which are only partially characterized. The reported test-set metrics (RMSE 66 K, F1=0.98) do not directly transfer to production, where CMD-based contamination is ~19% and only 57.1% of visually inspected new candidates are confirmed.

major comments (4)
  1. [Section 4.1 / 4.1.2 / 4.1.3] The reported e_Teff is a pure random-uncertainty estimate (Section 4.1, MC propagation of flux errors). The comparison against genuinely independent PMS high-resolution studies in Section 4.1.2 shows weighted RMS scatter up to ~241 K and systematic offsets of ~100–180 K for spotted components, comparable to or larger than the claimed precision. The PHOENIX calibration (Section 2.1) is the single load-bearing anchor, and Section 4.1.3 explicitly acknowledges this model dependence but does not quantify it. The catalog should report a systematic error term alongside e_Teff, or at least provide a clear statement that e_Teff is only a random-error bar and that the total uncertainty is likely several times larger. Without this, quoting 'RMSE ~66 K' as the precision is misleading for users of the 1.73-million-row catalog.
  2. [Section 4.2, Figure 5 (bottom panels)] The production probability threshold p_thr is selected using a circular procedure: the gLiI feature is both an input to the classifier and the variable used to decide the threshold. Specifically, 'N_Tot = N(gLiI>=0.2)+N(gLiI<0.2)' and the threshold is chosen when these two terms become 'comparable'/'subdominant'. The criterion is not defined quantitatively, and no justification is given for why this yields a decision boundary that controls false positives. This choice directly determines the final 3,121 TTS candidate spectra. The authors should either adopt a pre-registered threshold based on labeled data (e.g., a target false-positive rate) or demonstrate on a held-out set that the gLiI-based threshold behaves as claimed.
  3. [Section 4.2.1] The production reliability of the TTS classifier is much weaker than the test-set metrics suggest. The CMD contamination is 19.3%, and only 588 of 1,029 visually inspected previously unreported candidates (57.1%) are confirmed as TTS. The paper acknowledges these issues, but the abstract and conclusions emphasize the 2,534-candidate list without prominently stating that a large fraction may be false positives. For a catalog paper, the recommended usage should be clarified: users should be told that the candidate list has ~20% contamination and that a higher probability threshold or additional filters (e.g., CMD criterion) can produce a cleaner subsample. The current framing overstates the actionable precision of the TTS catalog.
  4. [Section 2.1 / 4.1.1 / 5] The validation against LASP, GSP-Phot, APOGEE-Net, and BOSS-Net is not an independent check of the PHOENIX temperature scale, because all of these methods ultimately rely on synthetic-atmosphere models or training labels with similar assumptions. The conclusion in Section 5 that the model 'effectively maps synthetic spectral indices to their empirical counterparts' (r_S=0.95) overstates the evidence. The only partially independent validations are the PMS-focused comparisons in Section 4.1.2, which show significant scatter and offsets, and the old-cluster contamination check, which is a classification test rather than a Teff test. The authors should temper the wording and, ideally, include a comparison against a small sample of stars with fundamental Teff determinations (e.g., interferometric radii or eclipsing binaries) to place a floor on the systematic uncertainty.
minor comments (6)
  1. [Abstract] The abstract states 'Using nine spectral features as inputs, we train a Gradient Boosting Machine...' but Section 4.1 uses seven features for the GBM and Section 4.2 uses eight for the LR classifier. Please reconcile the count.
  2. [Abstract / Section 4.1] The total number of LMS spectra is given as 1,733,802 in the abstract and 1,733,852 in Section 4.1. Correct the inconsistency.
  3. [Section 4.2 (CTTS/WTTS criterion)] The polynomial for the CTTS/WTTS boundary uses log(Teff) without specifying the base. State explicitly that it is base-10, as implied by the figure axis and the coefficients.
  4. [Figure 2 (top)] The learning curve is described as showing 'approximately 100 trees represent the optimal number', but the curve is labeled for 130 trees. Clarify whether the selected 100-tree model is the minimum of the validation curve or a separate choice.
  5. [Section 2.3 / Table A1] The feature list in Table A1 is long and many indices are unused. A concise table of only the used features in the main text would improve readability, with the full table relegated to the appendix (as already done).
  6. [General] Several passages are verbose and could be shortened to improve readability — for example, the descriptions of logistic regression and GBM in Sections 3.1–3.2 contain textbook-level detail that could be condensed.

Circularity Check

0 steps flagged

No significant circularity: Teff is anchored to the PHOENIX synthetic grid and TTS labels are external; self-references are not load-bearing.

full rationale

The central Teff regression is trained on PHOENIX synthetic spectra with known Teff labels (Section 2.1) and applied to LAMOST EW features; no parameter is fitted to the validation catalogs, so the comparison with LASP, Gaia/GSP-Phot, APOGEE-Net, and BOSS-Net is an external check rather than a fitted prediction. The TTS classifier is supervised on bona fide Taurus/OSFC members and Gaia golden-sample non-members (Section 2.2.1), and its candidates are tested against independent catalogs (McBride et al., Saad et al., Fang et al., Olivares et al., Zhang et al.) and old-cluster samples. Several cited works share co-authors (Hernandez et al. 2004; Sanchez-Sanjuan et al. 2024; Munoz et al. 2026), but they supply measurement conventions and a kinematic candidate catalog, not a uniqueness theorem or the target predictions themselves, so they are not load-bearing circularity. The acknowledged PHOENIX model dependence and the lack of a fundamental temperature-scale validation (Section 4.1.3) are accuracy/robustness caveats, not definitional or fitted-input circularity.

Axiom & Free-Parameter Ledger

4 free parameters · 6 axioms · 0 invented entities

The paper's contribution rests on external synthetic atmosphere models and assumed-representative labeled samples; no physical constants are fit to data. The main free choices are the GBM hyperparameters, the kernel-PCA dimension, and the production probability threshold.

free parameters (4)
  • GBM hyperparameters (n_trees, learning_rate, max_depth) = 100 / 0.1 / 5
    Selected by grid search + 5-fold cross-validation on the synthetic sample; not physically motivated.
  • Number of kernel-PCA components = 2 (92.3% variance)
    Chosen as hyperparameter in LR training; no physical basis.
  • Production probability threshold p_thr = base 0.1; distance-bin specific p_thr_d
    Set so that N(gLiI<0.2) becomes subdominant in each 0.1 kpc bin, effectively tuning the threshold on the production data.
  • gLiI missing-value imputation = 0.0
    Missing Gaussian-fit Li I EW is set to zero rather than modeled; affects classifier inputs.
axioms (6)
  • domain assumption PHOENIX synthetic spectra (solar metallicity, Teff 2300–6000 K, log g 3.0–5.0, A_V 0–5, S/N 30–500) adequately represent real LAMOST low-mass spectra for the purpose of Teff calibration.
    The GBM regressor is trained exclusively on this synthetic library without observed labels; if the synthetic line strengths or temperature scale are biased, all Teff are biased (Sec. 2.1, 4.1.3).
  • domain assumption The bandpass spectral-index measurements (TiO, VO, H-alpha, Li I) are robust to reddening, S/N and continuum calibration as claimed in Hernandez et al. (2004).
    The method and the derived uncertainties depend on this; only flux>3 sigma per band is used as quality cut (Sec. 2.3).
  • domain assumption Taurus and OSFC catalogs used as TTS labels are pure and representative of the TTS population in the LAMOST 1-kpc sample.
    552 TTS spectra from two star-forming regions define the positive class (Sec. 2.2.1); if they are incomplete or biased, the production classification inherits the bias.
  • domain assumption Gaia DR3 golden-sample FGKM stars used as non-TTS are not contaminated by young stars or strong activity.
    The negative class is drawn from these catalogs with LAMOST subclass matching; active binaries could mimic TTS features (Sec. 2.2.1).
  • domain assumption The CMD cut (RP - 5log(1000/pi)+5 < 2.1(BP-RP)+3.0) separates PMS sources from old field stars as intended.
    Used to estimate the 19.3% contamination rate; relies on the PARSEC 50-Myr isochrone and a linear approximation (Sec. 4.2.1).
  • standard math Standard statistical assumptions of supervised ML (i.i.d. labeled samples, no leakage, correct train/test splits) hold.
    The labeled sample is split 80/20 and the production sample excludes labeled spectra; assumptions underpin the reported F1/Brier scores.

pith-pipeline@v1.3.0-alltime-deepseek · 28390 in / 15768 out tokens · 151955 ms · 2026-08-01T20:31:00.639644+00:00 · methodology

0 comments
read the original abstract

We present effective temperature (Teff) estimates of low-mass stars and T Tauri stars (TTS) candidate detections derived from automated spectroscopic measurements and low-complexity machine-learning models applied to the LAMOST DR10 V2 survey. Equivalent widths of key diagnostic spectral features, including the atomic lines Halpha, LiI 6708 AA and TiO/VO molecular bands, are automatically measured for all stars within 1 kpc observed by LAMOST. Using nine spectral features as inputs, we train a Gradient Boosting Machine tree-based regression model, calibrated with synthetic spectra from the PHOENIX library, to predict Teff over the range 2,500 - 5,100 K. We apply a logistic regression model to the principal components derived from the measured spectral features, enabling efficient identification of TTS candidates. Both models exhibit strong performance in validation tests. Finally, a Monte Carlo framework is employed to propagate input uncertainties and estimate Teff and its associated uncertainties for low-mass stars from 1,733,802 spectra and to identify 2,534 candidate TTS from 3,121 spectra.

Figures

Figures reproduced from arXiv: 2607.16585 by B. Sabogal, C. D. Millan-Valderrama, J. Hernandez, J. Mu\~noz.

Figure 1
Figure 1. Figure 1: Correlation matrix for the 18 spectral features calculated over the 1309 sources with full measurements (485 TTS and 824 non-TTS). to the same molecular species is expected, as they are primarily sen￾sitive to temperature (Herczeg & Hillenbrand 2014). A similar ra￾tionale applies to the moderately strong correlation found between the features of the two different molecular species included in this analysis… view at source ↗
Figure 2
Figure 2. Figure 2: Top: Learning curve for our GBM regression model to determine 𝑇𝑒 𝑓 𝑓 . Bottom: Feature importance of our GBM regression model through of the Mean SHAP values. operates in two stages. First, it performs a coarse search over a grid of synthetic and empirical templates to obtain initial estimates of 𝑇𝑒 𝑓 𝑓 , surface gravity, metallicity, and radial velocity. In the second stage, these parameters are refined t… view at source ↗
Figure 3
Figure 3. Figure 3: Comparison of the 𝑇𝑒 𝑓 𝑓 within 0.4 kpc for around 6,300 common sources of the catalogs LAMOST (top left panel), Gaia (top right panel), APOGEE (bottom left panel) (Olney et al. 2020), and BOSS (bottom right panel) (Sizemore et al. 2024) with these obtained by our GBM regression model. The Spearman correlation coefficient is indicated in each inset. catalogs with known stellar parameters, allowing the mode… view at source ↗
Figure 4
Figure 4. Figure 4: Comparison between 𝑇𝑒 𝑓 𝑓 derived with our GBM regressor and values reported in dedicated studies of PMS stars. Grey circles correspond to Frasca et al. (2025), red diamonds to Gangi et al. (2022), and blue symbols to Carini et al. (2026), using ESPRESSO (triangles) and X-Shooter (squares) spectra. Red open squares indicate 𝑇𝑒 𝑓 𝑓 reported by Gangi et al. (2022) us￾ing a two-temperature approach, correspon… view at source ↗
Figure 5
Figure 5. Figure 5: Top left: Model stability assessment: cumulative mean F1-score with ± 1 standard deviation band over 300 random training/validation splits. Top right: Confusion matrix for test sample for our LR classification model. With only 3 false positives and 2 false negatives among the 275 test samples, the model demonstrates a balanced and robust classification performance for both TTS and non-TTS objects. Bottom l… view at source ↗
Figure 6
Figure 6. Figure 6: Separation boundary between CTTS and WTTS. This figure is sim￾ilar to the Saad et al. (2024) that stablishes a limit through H𝛼 as a function of log(𝑇𝑒 𝑓 𝑓 ) for separate the star populations CTTS and WTTS. caused by moonlight scattering on bright nights could affect the ac￾curacy of the sky-subtraction during the LAMOST reduction process (e.g., Bai et al. 2017). This issue can also occur in star-forming r… view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

117 extracted references · 17 canonical work pages

  1. [1]

    Deep Learning , author=

  2. [2]

    2006 , publisher =

    Pattern Recognition and Machine Learning , author =. 2006 , publisher =

    2006

  3. [3]

    2023 , issn =

    Leakage and the reproducibility crisis in machine-learning-based science , journal =. 2023 , issn =. doi:https://doi.org/10.1016/j.patter.2023.100804 , author =

    arXiv 2023

  4. [4]

    , keywords =

    Rotational Velocities and Chromospheric/Coronal Activity of Low-Mass Stars in the Young Open Clusters IC 2391 and IC 2602. , keywords =. doi:10.1086/303930 , adsurl =

    work page doi:10.1086/303930

  5. [5]

    X-shooter study of accretion in Chamaeleon I. II. A steeper increase of accretion with stellar mass for very low-mass stars?. , keywords =. doi:10.1051/0004-6361/201630147 , archivePrefix =. 1704.02842 , primaryClass =

    Pith/arXiv arXiv

  6. [6]

    X-shooter spectroscopy of young stellar objects. II. Impact of chromospheric emission on accretion rate estimates. , keywords =. doi:10.1051/0004-6361/201220921 , archivePrefix =. 1301.3058 , primaryClass =

    Pith/arXiv arXiv

  7. [7]

    , keywords =

    Characterizing the IYJ Excess Continuum Emission in T Tauri Stars. , keywords =. doi:10.1088/0004-637X/730/2/73 , archivePrefix =. 1101.4649 , primaryClass =

    Pith/arXiv arXiv

  8. [8]

    , keywords =

    The Effects of Starspots on Spectroscopic Mass Estimates of Low-mass Young Stars. , keywords =. doi:10.3847/1538-4357/ac37bd , archivePrefix =. 2111.03957 , primaryClass =

    Pith/arXiv arXiv

  9. [9]

    , keywords =

    The Gaia-ESO Survey: Chromospheric emission, accretion properties, and rotation in Velorum and Chamaeleon I. , keywords =. doi:10.1051/0004-6361/201424409 , archivePrefix =. 1412.4153 , primaryClass =

    Pith/arXiv arXiv

  10. [10]

    PENELLOPE: The ESO data legacy program to complement the Hubble UV Legacy Library of Young Stars (ULLYSES). I. Survey presentation and accretion properties of Orion OB1 and -Orionis. , keywords =. doi:10.1051/0004-6361/202140639 , archivePrefix =. 2103.12446 , primaryClass =

    Pith/arXiv arXiv

  11. [11]

    , keywords =

    The Effect of Starspots on Spectroscopic Age and Mass Estimates of Nonaccreting T Tauri Stars in the Taurus─Auriga Star-forming Region. , keywords =. doi:10.3847/1538-4357/ad393b , archivePrefix =. 2403.20255 , primaryClass =

    Pith/arXiv arXiv

  12. [12]

    Lithium, iron, and barium elemental abundances in eight nearby young clusters

    PENELLOPE: IX. Lithium, iron, and barium elemental abundances in eight nearby young clusters. , keywords =. doi:10.1051/0004-6361/202659572 , archivePrefix =. 2604.02919 , primaryClass =

    Pith/arXiv arXiv

  13. [13]

    , keywords =

    LAMOST medium-resolution observations of the Pleiades. , keywords =. doi:10.1051/0004-6361/202553673 , archivePrefix =. 2504.09031 , primaryClass =

    Pith/arXiv arXiv

  14. [14]

    GIARPS High-resolution Observations of T Tauri stars (GHOsT). IV. Accretion properties of the Taurus-Auriga young association. , keywords =. doi:10.1051/0004-6361/202244042 , archivePrefix =. 2208.14895 , primaryClass =

    Pith/arXiv arXiv

  15. [15]

    , keywords =

    Spectral Biases, Starspot Morphology, and Dynamo Transitions on the Pre-main Sequence: Insights from the X-Shooter WTTS Library. , keywords =. doi:10.3847/1538-4357/adf6ad , archivePrefix =. 2505.10837 , primaryClass =

    Pith/arXiv arXiv

  16. [16]

    , keywords =

    A New Generation of PARSEC-COLIBRI Stellar Isochrones Including the TP-AGB Phase. , keywords =. doi:10.3847/1538-4357/835/1/77 , archivePrefix =. 1701.08510 , primaryClass =

    Pith/arXiv arXiv

  17. [17]

    , keywords =

    A Catalog of Young Stellar Objects from the LAMOST and ZTF Surveys. , keywords =. doi:10.3847/1538-4365/acd84b , adsurl =

    work page doi:10.3847/1538-4365/acd84b

  18. [18]

    Hosmer, D. W. and Lemeshow, S. , year =. A goodness-of-fit test for the multiple logistic regression model , journal =

  19. [19]

    , keywords =

    An Efficient Framework for Automatic Carbon Star Detection Using Machine Learning Techniques on LAMOST DR9 and Gaia DR3. , keywords =. doi:10.1088/1538-3873/ae3f57 , adsurl =

    doi:10.1088/1538-3873/ae3f57

  20. [20]

    The CDS reference database for astronomical objects

    The SIMBAD astronomical database. The CDS reference database for astronomical objects. , keywords =. doi:10.1051/aas:2000332 , archivePrefix =. astro-ph/0002110 , primaryClass =

    Pith/arXiv arXiv

  21. [21]

    , keywords =

    A novel approach for variable star classification based on imbalanced learning. , keywords =. doi:10.1017/pasa.2023.35 , adsurl =

    work page doi:10.1017/pasa.2023.35 2023

  22. [22]

    , keywords =

    LAMOST Spectrograph Response Curves: Stability and Application to Flux Calibration. , keywords =. doi:10.3847/1538-4365/227/2/27 , archivePrefix =. 1611.08216 , primaryClass =

    Pith/arXiv arXiv

  23. [23]

    , keywords =

    LAMOST meets Gaia: The Galactic open clusters. , keywords =. doi:10.1051/0004-6361/202243590 , archivePrefix =. 2207.09121 , primaryClass =

    Pith/arXiv arXiv

  24. [24]

    RAA , keywords =

    Sky subtraction for LAMOST. RAA , keywords =. doi:10.1088/1674-4527/17/9/91 , archivePrefix =. 1705.02079 , primaryClass =

    Pith/arXiv arXiv

  25. [25]

    , keywords =

    A Sky-Subtraction Algorithm for LAMOST Using Two-Dimensional Sky-Background Modeling. , keywords =. doi:10.1071/AS11071 , adsurl =

    work page doi:10.1071/as11071

  26. [26]

    Shen, Xinwei and Meinshausen, Nicolai , title =. J. R. Stat. Soc. Ser. B Stat. Methodol. , volume =. 2024 , month =. doi:10.1093/jrsssb/qkae108 , url =

    doi:10.1093/jrsssb/qkae108 2024

  27. [27]

    arXiv e-prints , keywords =

    Machine Learning in Astronomy: a practical overview. arXiv e-prints , keywords =. doi:10.48550/arXiv.1904.07248 , archivePrefix =. 1904.07248 , primaryClass =

    Pith/arXiv arXiv doi:10.48550/arxiv.1904.07248 1904

  28. [28]

    , keywords =

    A Self-consistent Data-driven Model for Determining Stellar Parameters from Optical and Near-infrared Spectra. , keywords =. doi:10.3847/1538-3881/ad291d , archivePrefix =. 2402.05184 , primaryClass =

    Pith/arXiv arXiv

  29. [29]

    , keywords =

    APOGEE Net: An Expanded Spectral Model of Both Low-mass and High-mass Stars. , keywords =. doi:10.3847/1538-3881/ac4de7 , archivePrefix =. 2201.03661 , primaryClass =

    Pith/arXiv arXiv

  30. [30]

    , keywords =

    Faint disks around classical T Tauri stars: Small but dense enough to form planets. , keywords =. doi:10.1051/0004-6361/201322388 , archivePrefix =. 1402.5312 , primaryClass =

    Pith/arXiv arXiv

  31. [31]

    The dispersal of planet-forming discs: theory confronts observations. R. Soc. Open Sci. , keywords =. doi:10.1098/rsos.170114 , archivePrefix =. 1704.00214 , primaryClass =

    Pith/arXiv arXiv

  32. [32]

    Third Catalog of Emission-Line Stars of the Orion Population : 3 : 1988

    1988

  33. [33]

    , keywords =

    Very Low Mass Stars and Brown Dwarfs in Taurus-Auriga. , keywords =. doi:10.1086/344673 , archivePrefix =. astro-ph/0209164 , primaryClass =

    Pith/arXiv arXiv

  34. [34]

    , keywords =

    Measuring the Magnetic Field on the Classical T Tauri Star BP Tauri. , keywords =. doi:10.1086/307128 , adsurl =

    work page doi:10.1086/307128

  35. [35]

    , year = 2016, month = sep, volume =

    Accretion onto Pre-Main-Sequence Stars. , year = 2016, month = sep, volume =. doi:10.1146/annurev-astro-081915-023347 , adsurl =

    doi:10.1146/annurev-astro-081915-023347 2016

  36. [36]

    Proceedings of the International Astronomical Union , volume=

    Magnetic fields of weak line T-Tauri stars , author=. Proceedings of the International Astronomical Union , volume=. 2016 , publisher=

    2016

  37. [37]

    The CIDA Variability Survey of Orion OB1. II. Demographics of the Young, Low-mass Stellar Populations. , keywords =. doi:10.3847/1538-3881/aaf79b , adsurl =

    doi:10.3847/1538-3881/aaf79b

  38. [38]

    A Census of the Low Accretors. I. The Catalog. , keywords =. doi:10.3847/1538-3881/ac3ee6 , archivePrefix =. 2201.07707 , primaryClass =

    Pith/arXiv arXiv

  39. [39]

    Signs of accretion and coronal activity , author=

    X-ray and UV radiation in the planet-forming T-Tauri system PDS 70. Signs of accretion and coronal activity , author=. MNRAS , volume=. 2023 , publisher=

    2023

  40. [40]

    Astronomy Letters , volume=

    T Tauri stars: Physical parameters and evolutionary status , author=. Astronomy Letters , volume=. 2016 , publisher=

    2016

  41. [41]

    ABYSS. II. Identification of Young Stars in Optical SDSS Spectra and Their Properties. , keywords =. doi:10.3847/1538-3881/ad2001 , archivePrefix =. 2401.01932 , primaryClass =

    Pith/arXiv arXiv

  42. [42]

    An Optical Spectroscopic Study of T Tauri Stars. I. Photospheric Properties. , keywords =. doi:10.1088/0004-637X/786/2/97 , archivePrefix =. 1403.1675 , primaryClass =

    Pith/arXiv arXiv

  43. [43]

    , keywords =

    Pre-Main-Sequence Evolution in the Taurus-Auriga Molecular Cloud. , keywords =. doi:10.1086/192235 , adsurl =

    doi:10.1086/192235

  44. [44]

    Data Science Journal , volume=

    Astronomy in the big data era , author=. Data Science Journal , volume=

  45. [45]

    RAA , year = 2012, month = sep, volume =

    Data release of the LAMOST pilot survey. RAA , year = 2012, month = sep, volume =. doi:10.1088/1674-4527/12/9/004 , adsurl =

    doi:10.1088/1674-4527/12/9/004 2012

  46. [46]

    RAA , keywords =

    LASPM: the LAMOST stellar parameter pipeline for M-type stars and application to the sixth and seventh data release (DR6 and DR7). RAA , keywords =. doi:10.1088/1674-4527/21/8/202 , adsurl =

    doi:10.1088/1674-4527/21/8/202

  47. [47]

    , keywords =

    A LAMOST Spectroscopic Study of T Tauri Stars in the Orion OB1a Subassociation. , keywords =. doi:10.3847/1538-3881/acc467 , adsurl =

    work page doi:10.3847/1538-3881/acc467

  48. [48]

    X-shooter spectroscopy of young stellar objects. IV. Accretion in low-mass stars and substellar objects in Lupus. , keywords =. doi:10.1051/0004-6361/201322254 , archivePrefix =. 1310.2069 , primaryClass =

    Pith/arXiv arXiv 2069

  49. [49]

    Big data in astronomy: from evolution to revolution , author=. Int. J. Adv. Astron. , volume=

  50. [50]

    arXiv preprint arXiv:2502.15300 , year=

    Machine Learning in Stellar Astronomy: Progress up to 2024 , author=. arXiv preprint arXiv:2502.15300 , year=

    Pith/arXiv arXiv 2024

  51. [51]

    Summary of the contents and survey properties

    Gaia Data Release 2. Summary of the contents and survey properties. , keywords =. doi:10.1051/0004-6361/201833051 , archivePrefix =. 1804.09365 , primaryClass =

    Pith/arXiv arXiv

  52. [52]

    2MASS All Sky Catalog of point sources

  53. [53]

    MNRAS , volume =

    Ramachandra, Nesar and Chaves-Montero, Jonás and Alarcon, Alex and Fadikar, Arindam and Habib, Salman and Heitmann, Katrin , title =. MNRAS , volume =. 2022 , month =. doi:10.1093/mnras/stac1790 , url =

    work page doi:10.1093/mnras/stac1790 2022

  54. [54]

    2021 , publisher=

    The observation and analysis of stellar photospheres , author=. 2021 , publisher=

    2021

  55. [55]

    Dwarfs and subgiants

    An absolutely calibrated T _ eff scale from the infrared flux method. Dwarfs and subgiants. , keywords =. doi:10.1051/0004-6361/200913204 , archivePrefix =. 1001.3142 , primaryClass =

    Pith/arXiv arXiv

  56. [56]

    2019 , publisher=

    Advanced guide to Python 3 programming , author=. 2019 , publisher=

    2019

  57. [57]

    Herbig Ae/Be Stars: An Investigation of Molecular Environments and Associated Stellar Populations

  58. [58]

    , keywords =

    Spectral Analysis and Classification of Herbig Ae/Be Stars. , keywords =. doi:10.1086/381908 , adsurl =

    work page doi:10.1086/381908

  59. [59]

    , year = 1976, month = dec, volume =

    Error propagation by the Monte Carlo method in geochemical calculations. , year = 1976, month = dec, volume =. doi:10.1016/0016-7037(76)90092-2 , adsurl =

    doi:10.1016/0016-7037(76)90092-2 1976

  60. [60]

    , keywords =

    A Survey for New Members of Taurus from Stellar to Planetary Masses. , keywords =. doi:10.3847/1538-3881/ab2594 , archivePrefix =. 1907.00055 , primaryClass =

    Pith/arXiv arXiv 1907

  61. [61]

    RAA , year = 2012, month = sep, volume =

    The Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST). RAA , year = 2012, month = sep, volume =. doi:10.1088/1674-4527/12/9/003 , adsurl =

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

  62. [62]

    , keywords =

    A classifier for spurious astrometric solutions in Gaia eDR3. , keywords =. doi:10.1093/mnras/stab3588 , archivePrefix =. 2101.11641 , primaryClass =

    Pith/arXiv arXiv

  63. [63]

    Parallax bias versus magnitude, colour, and position

    Gaia Early Data Release 3. Parallax bias versus magnitude, colour, and position. , keywords =. doi:10.1051/0004-6361/202039653 , archivePrefix =. 2012.01742 , primaryClass =

    Pith/arXiv arXiv 2012

  64. [64]

    The astrometric solution

    Gaia Early Data Release 3. The astrometric solution. , keywords =. doi:10.1051/0004-6361/202039709 , archivePrefix =. 2012.03380 , primaryClass =

    Pith/arXiv arXiv 2012

  65. [65]

    A golden sample of astrophysical parameters

    Gaia Data Release 3. A golden sample of astrophysical parameters. , keywords =. doi:10.1051/0004-6361/202243800 , archivePrefix =. 2206.05870 , primaryClass =

    Pith/arXiv arXiv

  66. [66]

    , keywords =

    A new extensive library of PHOENIX stellar atmospheres and synthetic spectra. , keywords =. doi:10.1051/0004-6361/201219058 , archivePrefix =. 1303.5632 , primaryClass =

    Pith/arXiv arXiv

  67. [67]

    Model Atmospheres for M (Sub)Dwarf Stars. I. The Base Model Grid. , keywords =. doi:10.1086/175708 , archivePrefix =. astro-ph/9601150 , primaryClass =

    Pith/arXiv arXiv

  68. [68]

    , keywords =

    Kinematic study of the Orion Complex: analysing the young stellar clusters from big and small structures. , keywords =. doi:10.1093/mnras/stae2157 , archivePrefix =. 2409.09206 , primaryClass =

    Pith/arXiv arXiv

  69. [69]

    Advanced Software, Control, and Communication Systems for Astronomy , year = 2004, editor =

    Design and implementation of the spectra reduction and analysis software for LAMOST Telescope. Advanced Software, Control, and Communication Systems for Astronomy , year = 2004, editor =. doi:10.1117/12.548737 , adsurl =

    work page doi:10.1117/12.548737 2004

  70. [70]

    , keywords =

    The Relationship between Infrared, Optical, and Ultraviolet Extinction. , keywords =. doi:10.1086/167900 , adsurl =

    doi:10.1086/167900

  71. [71]

    , keywords =

    On the Construction of a New Stellar Classification Template Library for the LAMOST Spectral Analysis Pipeline. , keywords =. doi:10.1088/0004-6256/147/5/101 , adsurl =

    work page doi:10.1088/0004-6256/147/5/101

  72. [72]

    Logistic regression was as good as machine learning for predicting major chronic diseases , journal=

    Simon Nusinovici and Yih Chung Tham and Marco Yu. Logistic regression was as good as machine learning for predicting major chronic diseases , journal=. 2020 , issn =. doi:https://doi.org/10.1016/j.jclinepi.2020.03.002 , url =

    work page doi:10.1016/j.jclinepi.2020.03.002 2020

  73. [73]

    Collins and Ewout W

    Evangelia Christodoulou and Jie Ma and Gary S. Collins and Ewout W. Steyerberg and Jan Y. Verbakel and Ben. A systematic review shows no performance benefit of machine learning over logistic regression for clinical prediction models , journal =. 2019 , issn =. doi:https://doi.org/10.1016/j.jclinepi.2019.02.004 , url =

    doi:10.1016/j.jclinepi.2019.02.004 2019

  74. [74]

    Advances in Data Sciences, Security and Applications: Proceedings of ICDSSA 2019 , pages=

    Classification of Astronomical Objects Using Various Machine Learning Techniques , author=. Advances in Data Sciences, Security and Applications: Proceedings of ICDSSA 2019 , pages=. 2020 , doi=

    2019

  75. [75]

    Analyzing astronomical data with machine learning techniques , author=. Astron. Astrophys. Trans. , volume=. 2022 , doi=

    2022

  76. [76]

    arXiv preprint arXiv:2302.11573 , year=

    Analyzing astronomical data with machine learning techniques , author=. arXiv preprint arXiv:2302.11573 , year=

    Pith/arXiv arXiv

  77. [77]

    An application to colour-colour diagrams

    On the use of logistic regression for stellar classification. An application to colour-colour diagrams. Exp. Astron. , keywords =. doi:10.1007/s10686-018-9591-4 , archivePrefix =. 1805.09551 , primaryClass =

    Pith/arXiv arXiv

  78. [78]

    , keywords =

    Logistic regression to boost exoplanet detection performances. , keywords =. doi:10.1093/mnras/stae2657 , archivePrefix =. 2403.11571 , primaryClass =

    Pith/arXiv arXiv

  79. [79]

    , title =

    Stoltzfus, Jill C. , title =. Acad. Emerg. Med. , volume =. doi:https://doi.org/10.1111/j.1553-2712.2011.01185.x , url =. https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1553-2712.2011.01185.x , abstract =

    arXiv 2011

  80. [80]

    2019 , publisher=

    Robust statistics: theory and methods (with R) , author=. 2019 , publisher=

    2019

Showing first 80 references.