Pith. sign in

REVIEW 3 major objections 5 minor 99 references

MAISTEP -- a new grid-based machine learning tool for inferring stellar parameters I. Ages of giant-planet host stars

T0 review · 3 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a stacked ensemble of four tree-based machine learners, trained on a grid of stellar models and fed only effective temperature, metallicity, and luminosity, recovers stellar radius, mass, and age to within a few…

desk verdict A solid ML tool for radius and mass with credible seismic validation, but the headline Hot Jupiter youth signal is confounded by the sample's mass difference and isn't established. read the letter →

arxiv 2502.02176 v1 pith:VOVDPE3F submitted 2025-02-04 astro-ph.SR astro-ph.EPastro-ph.IM

Juma Kamulali , Benard Nsamba , Vardan Adibekyan , Achim Weiss , Tiago L. Campante , Nuno C. Santos This is my paper · ORCID
classification astro-ph.SRastro-ph.EPastro-ph.IM
keywords machinelearningstellaragesgiant-planethostshotJupitersgrid-basedinferencestackingparameters
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 develops a machine-learning tool, MAISTEP, that infers stellar radius, mass, and age from only atmosphere-level observables: effective temperature, [Fe/H], and a Gaia-based luminosity. It trains four tree-based regressors on a grid of stellar evolution models and combines them by weighted stacking. Against asteroseismic catalogues (APOKASC and LEGACY), radius and mass match to a few percent, while age scatter is 16-23 percent; the authors call these predictions commensurate with seismic inference. Applying the tool to 427 stars hosting Jupiter-mass planets yields characteristic ages of 1.98, 2.98, and 3.51 Gyr for Hot, Warm, and Cold Jupiter hosts, respectively, with the Hot-versus-colder differences statistically significant. The claim is that Hot Jupiters orbit statistically younger (and also more massive) main-sequence stars.

What carries the argument

The central object is a stacked ensemble of four tree-based regressors: Random Forest, Extra Trees, XGBoost, and CatBoost, trained on a grid of stellar models spanning masses $0.7$ to $1.6\,M_\odot$, metallicities $[{\rm Fe/H}] = -0.5$ to $0.5$ dex, and enrichment ratios $\Delta Y/\Delta Z = 0.4$ to $2.4$. The four base predictions are combined by non-negativity-constrained least squares, which assigns a weight to each algorithm; Extra Trees consistently receives the largest weight, and stacking reduces age bias by about 7% relative to the best single algorithm. Inputs are limited to $T_{\rm eff}$, $[{\rm Fe/H}]$, and $L$, making the tool applicable to the many exoplanet-host stars that lack asteroseismic measurements.

What would settle it

Construct a mass-matched control by resampling Warm and Cold Jupiter hosts to the same mass distribution as the Hot Jupiter hosts, then recompute the age distributions; if the 1.98, 2.98, and 3.51 Gyr differences collapse or invert, the central astrophysical claim is refuted.

Watch

Extended reading notes

Core claim

MAISTEP recovers main-sequence radius, mass, and age from $T_{\rm eff}$, [Fe/H], and a parallax-derived luminosity alone, without seismic data, at a precision the authors describe as commensurate with asteroseismic inference. Against APOKASC the bias and scatter are -0.5% (5%) in radius, 6% (5%) in mass, and -9% (16%) in age; against the LEGACY sample they are -0.2% (2%), -2% (3%), and 7% (23%), respectively. Applied to 427 stars hosting giant planets, the tool gives kernel-density age peaks of 1.98 Gyr for Hot Jupiter hosts, 2.98 Gyr for Warm Jupiter hosts, and 3.51 Gyr for Cold Jupiter hosts. Statistical tests reject identical distributions for Hot versus Warm and Hot versus Cold, while Warm versus Cold remains similar except in one tail-sensitive test. The paper concludes that Hot Jupiters are preferentially hosted by younger and more massive stars, corroborating earlier isochrone-based and kinematic results.

Load-bearing premise

The claim that Hot Jupiter hosts are intrinsically younger rests on the assumption that the age difference is not a side effect of the samples' differing stellar masses; the paper itself notes Hot Jupiter hosts are more massive, and no mass-matched comparison is presented.

Editorial extensions

If this is right

  • Stars without seismic data, which are the large majority of known exoplanet hosts, can be assigned radii, masses, and ages to few-percent (radius and mass) and roughly 20 percent (age) accuracy from spectroscopic and parallax data alone.
  • The inferred age distribution of giant-planet hosts supports the tidal-decay picture: close-in Jupiter-mass planets appear preferentially around younger stars because they spiral inward and are eventually engulfed on main-sequence timescales.
  • The Hot-Jupiter youth signal remains unchanged when the orbital-period cutoff defining Hot Jupiters is varied from 5 to 15 days, indicating the result is not an artifact of a specific boundary.
  • Because the training data are stellar models, the same pipeline can be used to test how changes in assumed model physics propagate into inferred ages for real planet-host ensembles.
  • The authors plan to extend the tool to subgiant and red-giant branches, where growing numbers of high-mass planet detections are being made.

Reading between the lines

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

  • The age gap between Hot and Warm/Cold Jupiter hosts could be partly a mass artifact: the paper's own mass distributions differ, with Hot Jupiter hosts peaking near $1.30\,M_\odot$ versus $1.09$ and $1.05\,M_\odot$ for the other two groups, and more massive main-sequence stars are younger, so a mass-matched resampling test is needed before attributing the youth signal to tidal physics.
  • The nearly perfect radius recovery is partly built into the inputs because luminosity and radius are tied through the Stefan-Boltzmann relation; the genuinely nontrivial inferences are mass and especially age, where scatter remains 16 to 23 percent.
  • The dominance of Extra Trees in the stacked weights suggests a simpler one- or two-algorithm variant might retain most of the predictive power, with the ensemble's main benefit appearing for age, the least constrained quantity.
  • The tool's validation is restricted to the main-sequence parameter space of the training grid; applying it to evolved hosts or stars with unusual helium enrichment or rotation histories would require grid extension and separate testing.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper presents MAISTEP, a stacking ensemble of four tree-based regressors (Random Forest, Extra Trees, XGBoost, CatBoost) trained on a MESA grid of main-sequence models (0.7-1.6 Msun, [Fe/H] -0.5 to 0.5, DeltaY/DeltaZ 0.4-2.4) to predict stellar radius, mass, and age from Teff, [Fe/H], and Gaia-based luminosity. The model is evaluated on held-out grid models and on 167 APOKASC and 55 LEGACY stars with seismic reference values, reporting radius biases below 0.5% with 2-5% scatter, mass biases -2% to 6% with 3-5% scatter, and age biases -9% to +7% with 16-23% scatter. The tool is then applied to 427 giant-planet host stars from the NASA Exoplanet Archive and SWEET-Cat; the authors report age distributions peaking at 1.98, 2.98, and 3.51 Gyr for Hot, Warm, and Cold Jupiter hosts and conclude that Hot Jupiter hosts are younger, confirming previous predictions.

Significance. If the central claims hold, MAISTEP would be a useful lightweight tool for estimating radii, masses, and ages of exoplanet host stars without seismic data, and the Hot Jupiter youth signal would strengthen evolutionary scenarios involving tidal inspiral. The validation design is non-circular: the models are trained on stellar-model outputs and tested against independent seismic catalogs, and the appendix compares against multiple pipelines (AIMS, BASTA, GOE). The radius and mass validation is credible, although the radius agreement is partly expected because Teff and L effectively fix R through the Stefan-Boltzmann relation, a point the authors themselves acknowledge. The age performance is weaker: biases of order 10%, 16-23% scatter, and median relative uncertainties of 23-28% are larger than the seismic reference uncertainties (9-17%). The strongest astrophysical conclusion is therefore the ensemble age comparison, and that comparison currently lacks a mass-matched control. Because the stated goal includes hypothesis-testing about the ages of giant-planet hosts, the mass confounding is the central issue to fix.

major comments (3)
  1. [Section 4, Fig. 7, Table 3] The claim that Hot Jupiter hosts are statistically younger than Warm/Cold Jupiter hosts is not established because the comparison is confounded by stellar mass. MAISTEP takes only Teff, [Fe/H], and L as inputs and learns age from the evolutionary-track mass-age relation, so at fixed atmospheric parameters a more massive main-sequence star is predicted to be younger. The bottom panel of Fig. 7 shows that the Hot Jupiter hosts have a peak mass of 1.30 Msun versus 1.09 and 1.05 Msun for Warm and Cold Jupiter hosts, and the mass rows of Table 3 show significant mass differences for categories A and B. The paper itself states that the mass difference 'may help explain the observed age differences.' The p-values in Table 3 test whether the unconditional age distributions differ; they do not test whether the age difference survives at fixed mass. A mass-matched or mass-stratified comparison (e.g., matching Hot Jupiter hosts to Warm/Cold hosts in mass bins, or including mass as a covariate or propensity score) is required before the age ranking can be attributed to the planet type rather than to the mass distributions.
  2. [Abstract and Section 5] The statement that the ML age predictions are 'commensurate with seismic inferences' goes beyond what the reported numbers support. The age comparisons show biases of -9% and +7%, scatters of 16% and 23%, and median relative uncertainties of 23% and 28%, whereas the APOKASC and AIMS reference ages have median relative uncertainties of 17% and 9%, respectively; the appendix comparison with the GOE pipeline gives a 20% bias and 32% scatter. These figures suggest that MAISTEP ages are useful for ensemble studies but not individually commensurate with asteroseismic ages. The conclusions do acknowledge that 'seismic information is needed to yield robust ages,' but the abstract's wording should be tempered or quantified, and the discussion should state explicitly that per-star ages are not competitive with seismic ages.
  3. [Section 4, Table 3] The significance tests in Table 3 are applied to point estimates (the median of each star's age distribution) and ignore the per-star age uncertainties, which the paper reports as 23-28% relative. Because the 10,000 Monte Carlo realizations are already computed for every star, the authors can propagate the full distributions into a hierarchical or bootstrap test; the current p-values should be interpreted as testing the point estimates only. This matters because the separation between the Hot and Cold Jupiter age peaks (1.98 vs 3.51 Gyr) is only about twice the typical per-star age uncertainty, and the point-estimate tests do not account for heteroscedastic measurement error or for stars counted in multiple planet categories.
minor comments (5)
  1. [Section 5, first paragraph] The age scatter for the APOKASC comparison is given as 12% here, whereas Section 3.2.2 and Fig. 5 report 16%; the abstract also says 16%. Please correct the inconsistency.
  2. [Abstract and Section 4] The abstract calls 1.98, 2.98, and 3.51 Gyr 'average age estimates,' but Section 4 describes them as peaks of the kernel-density distributions, with mean ages of 3.52, 4.41, and 5.13 Gyr. Please use 'peak' or 'mode' rather than 'average.'
  3. [Section 3.2.1, Eq. (4)] The luminosity uncertainties propagate only parallax and magnitude errors; the reddening correction from Eq. (5) uses E(B-V) with no associated uncertainty. A sensitivity test or an assumed uncertainty for E(B-V) would make the propagated age uncertainties more realistic.
  4. [Section 2.3] The text states that a maximum of 50 Optuna trials are used, but the hyperparameter search space is not described. A short description or table of the search ranges would improve reproducibility.
  5. [Table 2 caption] The caption uses 'def' without defining it in the notes; please clarify that it means the default value was used.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: ML training is independent of the test catalogs; the only self-referential element is a non-load-bearing companion-grid citation, and the Hot-Jupiter age result is a confounding limitation rather than a circular reduction.

full rationale

The central derivation chain is not circular. MAISTEP is trained on MESA stellar models to map Teff, [Fe/H], and L to radius, mass, and age; none of the target quantities is used as an input feature, and the stacking weights are fit on held-out cross-validation folds rather than on the quantities later called predictions. The tool is then tested against the external APOKASC and LEGACY seismic catalogs, so the validation does not reduce to the training set. The paper explicitly notes that L is non-independent of Teff and R through the Stefan-Boltzmann relation (Sec. 3.1), but this is a physical degeneracy, not a circular definition. The only self-citation is the pointer to Nsamba et al. (2025) for a detailed description of grid inputs (Sec. 2.1); the present paper itself specifies the grid ranges and input physics, so the citation is not load-bearing and does not force any result. Finally, the conclusion that Hot-Jupiter hosts are younger may be confounded by the fact that they are more massive, as the paper states: 'the Hot-Jupiter hosts are statistically more massive than stars hosting the Warm and Cold Jupiters, which may help explain the observed age differences' (Sec. 4, bottom panel of Fig. 7). No mass-matched analysis is presented, so the age ranking could arise from the model-inherited mass-age relation. That is a statistical and astrophysical correctness risk, not an equation-level circularity, and it does not undermine the external seismic validation of the tool.

Assumptions & free parameters 2 free parameters · 6 assumptions · 0 invented entities

The MAISTEP predictions rest on a grid of MESA models whose adopted physics (alpha_MLT, overshoot, diffusion, no rotation) is inherited from prior calibrations and a companion paper. The ML stacking adds fitted weights and tuned hyperparameters. No new particles, forces, or physical entities are introduced, and the only 'fitted' quantities are the standard machine-learning coefficients.

free parameters (2)
  • Stacking weights beta_j for each base algorithm and target (radius, mass, age) = XT receives the highest weights across all targets (Fig. 2)
    Fitted with non-negative least squares on cross-validation predictions from the training set; they define the final MAISTEP prediction.
  • Optuna-tuned hyperparameters (learning rate, max depth, n_estimators/iterations) for RF, XT, XGBoost, CatBoost = See Table 2
    Selected by minimizing RMSE on validation folds; they affect the accuracy and generalization of the base models.
assumptions (6)
  • domain assumption A single solar-calibrated mixing length, alpha_MLT = 1.71, applies to all grid models.
    Sec. 2.1: affects the structure and lifetime of every model, hence all inferred ages; no variation with mass, metallicity, or evolution is tested.
  • domain assumption Core overshoot is implemented with a fixed efficiency f_ov = 0.01 for convective cores.
    Sec. 2.1: overshoot extends main-sequence lifetime and thus biases ages if the real stars have different overshoot.
  • domain assumption Atomic diffusion is applied only to models with ZAMS mass below about 1.2 Msun, with only gravitational settling.
    Sec. 2.1: differential treatment changes surface [Fe/H] evolution and ages; the boundary is a modeling choice.
  • domain assumption Rotation is neglected because target stars are assumed to be slow rotators.
    Sec. 2.1: rotation alters mixing and lifetimes; the assumption is untested for the observed samples.
  • domain assumption The MESA stellar models with the listed reaction rates, opacities, EOS, and atmosphere are accurate representations of the real stars.
    Sec. 2.1 and Table 1: the ML training data inherit all model physics from this grid, described in detail in Nsamba et al. (2025).
  • domain assumption Gaia parallaxes, g-band magnitudes, and 3D dust map extinctions correctly determine the luminosities, and the seismic ages of APOKASC and LEGACY are valid benchmarks.
    Sec. 3.2.1 and 3.2.2: luminosity errors ignore extinction uncertainty, and the reference ages are themselves model-dependent.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MAISTEP -- a new grid-based machine learning tool for inferring stellar parameters I. Ages of giant-planet host stars." pith.science (2026). https://pith.science/paper/VOVDPE3F

@misc{pith2026250202176,
  author       = {Pith},
  title        = {Pith review of: MAISTEP -- a new grid-based machine learning tool for inferring stellar parameters I. Ages of giant-planet host stars},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VOVDPE3F}},
  note         = {Machine review of arXiv:2502.02176}
}
read the original abstract

Our understanding of exoplanet demographics partly depends on their corresponding host star parameters. With the majority of exoplanet-host stars having only atmospheric constraints available, robust inference of their parameters is susceptible to the approach used. The goal of this work is to develop a grid-based machine learning tool capable of determining the stellar radius, mass, and age using only atmospheric constraints and to analyse the age distribution of stars hosting giant planets. Our machine learning approach involves combining four tree-based machine learning algorithms (Random Forest, Extra Trees, Extreme Gradient Boosting, and CatBoost) trained on a grid of stellar models to infer stellar radius, mass, and age using Teff, [Fe/H], and luminosities. We perform a detailed statistical analysis to compare the inferences of our tool with those based on seismic data from the APOKASC and LEGACY samples. Finally, we apply our tool to determine the ages of stars hosting giant planets. Comparing the stellar parameter inferences from our machine learning tool with those from the APOKASC and LEGACY, we find a bias (and a scatter) of -0.5\% (5\%) and -0.2\% (2\%) in radius, 6\% (5\%) per cent and -2\% (3\%) in mass, and -9\% (16\%) and 7\% (23\%) in age, respectively. Therefore, our machine learning predictions are commensurate with seismic inferences. When applying our model to a sample of stars hosting Jupiter-mass planets, we find the average age estimates for the hosts of Hot Jupiters, Warm Jupiters, and Cold Jupiters to be 1.98, 2.98, and 3.51 Gyr, respectively. These statistical ages of the host stars confirm previous predictions - based on stellar model ages for a relatively small number of hosts, as well as on the average age-velocity dispersion relation - that stars hosting Hot Jupiters are statistically younger than those hosting Warm and Cold Jupiters.

Figures

Figures reproduced from arXiv: 2502.02176 by the authors.

Figure 1
Figure 1. Schematic representation of MAISTEP. The training data undergoes pre-processing before being passed to the different base algorithms under the model development phase, ultimately predicting the radius, mass, and age. See text for details [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Distribution of the weights (β) derived using nnls optimizer, which are employed to combine predictions of stellar radius (top panel), mass (middle panel), and age (bottom panel) from the base algorithms. See text for details [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Evaluation metric on the test sample as a function of the training size. The legend is organized in the order of performance, with the worst performers at the top and the best performers at the bottom. the algorithms required more samples (information) to explore in order to achieve adequate training. To test this, we trained our algorithms on the entire dataset and made predictions for some observed stars through s… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: HR diagram showing our selected sample, which include 167 stars from APOKASC catalogue plotted in green and 55 stars from the LEGACY sample shown in blue diamonds. In addition, we over-plot a few tracks at solar metallicity, [Fe/H] = 0, extending up to the bottom of th…
Figure 5
Figure 5. Figure 5: Each panel contains a one-to-one relation and fractional dif￾ferences from MAISTEP with respect to APOKASC as the reference values, for radius (top), mass (middle), and age (bottom). The black dashed lines in all panels represent the unity relation. The orange line in …
Figure 6
Figure 6. Figure 6: Same as [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: , we show the resulting age distributions of stellar popu￾lations hosting Hot Jupiters, Warm Jupiters, and Cold Jupiters, which are constructed solely from the median ages. The data show distinct peaks at ages of 1.98, 2.98, and 3.51 Gyr, with cor￾responding mean ages …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

99 extracted references · 65 canonical work pages

  1. [1]

    2019, in Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining, 2623–2631

    Akiba, T., Sano, S., Yanase, T., Ohta, T., & Koyama, M. 2019, in Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining, 2623–2631

    2019

  2. [2]

    Anderson, T. W. & Darling, D. A. 1954, Journal of the American statistical as- sociation, 49, 765

    1954

  3. [3]

    2018, Astronomy & Astrophysics, 616, A8

    Andrae, R., Fouesneau, M., Creevey, O., et al. 2018, Astronomy & Astrophysics, 616, A8

    2018

  4. [4]

    1999, Nuclear Physics A, 656, 3

    Angulo, C., Arnould, M., Rayet, M., et al. 1999, Nuclear Physics A, 656, 3

    1999

  5. [5]

    2015, Monthly Notices of the Royal Astronomical Society, 450, 1787

    Angus, R., Aigrain, S., Foreman-Mackey, D., & McQuillan, A. 2015, Monthly Notices of the Royal Astronomical Society, 450, 1787

    2015

  6. [6]

    2015, Astronomy & Astrophysics, 582, A25

    Appourchaux, T., Antia, H., Ball, W., et al. 2015, Astronomy & Astrophysics, 582, A25

    2015

  7. [7]

    J., & Scott, P

    Asplund, M., Grevesse, N., Sauval, A. J., & Scott, P. 2009, Annual review of astronomy and astrophysics, 47, 481

    2009

  8. [8]

    A., Weingrill, J., Fritzewski, D., Strassmeier, K

    Barnes, S. A., Weingrill, J., Fritzewski, D., Strassmeier, K. G., & Platais, I. 2016, The Astrophysical Journal, 823, 16

    2016

Show all 99 references
  1. [9]

    2019, arXiv preprint arXiv:1904.07248

    Baron, D. 2019, arXiv preprint arXiv:1904.07248

    2019 arXiv

  2. [10]

    2012, Monthly Notices of the Royal Astronomical Society, 427, 1847

    Bazot, M., Bourguignon, S., & Christensen-Dalsgaard, J. 2012, Monthly Notices of the Royal Astronomical Society, 427, 1847

    2012

  3. [11]

    P., Angelou, G

    Bellinger, E. P., Angelou, G. C., Hekker, S., et al. 2016, The Astrophysical Jour- nal, 830, 31

    2016

  4. [12]

    A., Huber, D., Gaidos, E., van Saders, J

    Berger, T. A., Huber, D., Gaidos, E., van Saders, J. L., & Weiss, L. M. 2020, The Astronomical Journal, 160, 108

    2020

  5. [13]

    Boyd, S. P. & Vandenberghe, L. 2004, Convex optimization (Cambridge univer- sity press)

    2004

  6. [14]

    1996, Machine learning, 24, 49

    Breiman, L. 1996, Machine learning, 24, 49

    1996

  7. [15]

    2001, Machine learning, 45, 5

    Breiman, L. 2001, Machine learning, 45, 5

    2001

  8. [16]

    2017, Classification and regression trees (Routledge) Article number, page 9 of 12 A&A proofs: manuscript no

    Breiman, L. 2017, Classification and regression trees (Routledge) Article number, page 9 of 12 A&A proofs: manuscript no. aanda

    2017

  9. [17]

    Buchhave, L. A. & Latham, D. W. 2015, The Astrophysical Journal, 808, 187

    2015

  10. [18]

    L., Barclay, T., Swift, J

    Campante, T. L., Barclay, T., Swift, J. J., et al. 2015, The Astrophysical Journal, 799, 170

    2015

  11. [19]

    L., Veras, D., North, T

    Campante, T. L., Veras, D., North, T. S., et al. 2017, Monthly Notices of the Royal Astronomical Society, 469, 1360

    2017

  12. [20]

    2014, Monthly Notices of the Royal Astronomical Society, 439, 2060

    Casagrande, L., Portinari, L., Glass, I., et al. 2014, Monthly Notices of the Royal Astronomical Society, 439, 2060

    2014

  13. [21]

    2013, The Astrophysical Journal Supple- ment Series, 210, 1

    Chaplin, W., Basu, S., Huber, D., et al. 2013, The Astrophysical Journal Supple- ment Series, 210, 1

    2013

  14. [22]

    2023, Proceedings of the National Academy of Sciences, 120, e2304179120

    Chen, D.-C., Xie, J.-W., Zhou, J.-L., et al. 2023, Proceedings of the National Academy of Sciences, 120, e2304179120

    2023

  15. [23]

    & Guestrin, C

    Chen, T. & Guestrin, C. 2016, in Proceedings of the 22nd acm sigkdd interna- tional conference on knowledge discovery and data mining, 785–794

    2016

  16. [24]

    R., van Saders, J

    Claytor, Z. R., van Saders, J. L., Santos, A. R., et al. 2020, The Astrophysical Journal, 888, 43

    2020

  17. [25]

    2015, Monthly Notices of the Royal Astronomical Society, 451, 3011

    Coelho, H., Chaplin, W., Basu, S., et al. 2015, Monthly Notices of the Royal Astronomical Society, 451, 3011

    2015

  18. [26]

    2022, Astronomy & As- trophysics

    Collaboration, G., Bailer-Jones, C., Teyssier, D., et al. 2022, Astronomy & As- trophysics

    2022

  19. [27]

    Cox, J. P. & Giuli, R. T. 1968, Principles of Stellar Structure: Physical Principles, V ol. 1 (Gordon and Breach)

    1968

  20. [28]

    H., Fields, B

    Cyburt, R. H., Fields, B. D., & Olive, K. A. 2003, Physics Letters B, 567, 227 Da Silva, R., de Mello, G. P., Milone, A., et al. 2012, Astronomy & Astrophysics, 542, A84

    2003

  21. [29]

    Dawson, R. I. & Johnson, J. A. 2018, Annual Review of Astronomy and Astro- physics, 56, 175

    2018

  22. [30]

    Deal, M., Cunha, M., Keszthelyi, Z., Perraut, K., & Holdsworth, D. L. 2021, Astronomy & Astrophysics, 650, A125 Des Etangs, A. L. & Lissauer, J. J. 2022, New Astronomy Reviews, 94, 101641

    2021

  23. [31]

    2021, The Astro- physical Journal, 920, 94

    Espinoza-Rojas, F., Chanamé, J., Jofré, P., & Casamiquela, L. 2021, The Astro- physical Journal, 920, 94

    2021

  24. [32]

    M., McMillan, P

    Feltzing, S., Howes, L. M., McMillan, P. J., & Stonkut ˙e, E. 2016, Monthly No- tices of the Royal Astronomical Society: Letters, slw209

    2016

  25. [33]

    W., Alexander, D

    Ferguson, J. W., Alexander, D. R., Allard, F., et al. 2005, The Astrophysical Journal, 623, 585

    2005

  26. [34]

    J., & Elsworth, Y

    Gai, N., Basu, S., Chaplin, W. J., & Elsworth, Y . 2011, The Astrophysical Jour- nal, 730, 63

    2011

  27. [35]

    2018, The Journal of Open Source Software, 3, 695

    Green, G. 2018, The Journal of Open Source Software, 3, 695

    2018

  28. [36]

    B., & Kallinger, T

    Gruberbauer, M., Guenther, D. B., & Kallinger, T. 2012, The Astrophysical Jour- nal, 749, 109

    2012

  29. [37]

    Hamer, J. H. & Schlaufman, K. C. 2019, The Astronomical Journal, 158, 190

    2019

  30. [38]

    Hara, N. C. & Ford, E. B. 2023, Annual Review of Statistics and Its Application, 10, 623

    2023

  31. [39]

    H., & Friedman, J

    Hastie, T., Tibshirani, R., Friedman, J. H., & Friedman, J. H. 2009, The ele- ments of statistical learning: data mining, inference, and prediction, V ol. 2 (Springer)

    2009

  32. [40]

    2000, arXiv preprint astro-ph/0007139

    Herwig, F. 2000, arXiv preprint astro-ph/0007139

    2000 arXiv

  33. [41]

    J., Christensen-Dalsgaard, J., et al

    Huber, D., Chaplin, W. J., Christensen-Dalsgaard, J., et al. 2013, The Astrophys- ical Journal, 767, 127

    2013

  34. [42]

    Iglesias, C. A. & Rogers, F. J. 1996, Astrophysical Journal v. 464, p. 943, 464, 943

    1996

  35. [43]

    2005, The European Physical Journal A-Hadrons and Nuclei, 25, 455

    Imbriani, G., Costantini, H., Formicola, A., et al. 2005, The European Physical Journal A-Hadrons and Nuclei, 25, 455

    2005

  36. [44]

    1933, Giorn

    Kolmogorov, A. 1933, Giorn. Ist. Ital. Attuari, 4 Krishna Swamy, K. 1966, Astrophysical Journal, vol. 145, p. 174, 145, 174

    1933

  37. [45]

    Lawson, C. L. & Hanson, R. J. 1995, Solving least squares problems (SIAM)

    1995

  38. [46]

    2018, Gaia Data Processing and Analysis Consortium, Tech

    Lindegren, L. 2018, Gaia Data Processing and Analysis Consortium, Tech. rep., GAIA-C3-TN-LU-LL-124

    2018

  39. [47]

    Liu, W. M. & Chaboyer, B. 2000, The Astrophysical Journal, 544, 818

    2000

  40. [48]

    N., Aguirre, V

    Lund, M. N., Aguirre, V . S., Davies, G. R., et al. 2017, The Astrophysical Jour- nal, 835, 172

    2017

  41. [49]

    Lund, M. N. & Reese, D. R. 2017, arXiv preprint arXiv:1711.01896

    2017 arXiv

  42. [50]

    2015, arXiv preprint arXiv:1510.06262

    Mamajek, E., Torres, G., Prsa, A., et al. 2015, arXiv preprint arXiv:1510.06262

    2015 arXiv

  43. [51]

    Mamajek, E. E. & Hillenbrand, L. A. 2008, The Astrophysical Journal, 687, 1264

    2008

  44. [52]

    & Queloz, D

    Mayor, M. & Queloz, D. 1995, nature, 378, 355

    1995

  45. [53]

    2010, Proceedings of the 9th Python in Science Conference

    McKinney, W., van der Walt, S., & Millman, J. 2010, Proceedings of the 9th Python in Science Conference

    2010

  46. [54]

    A., Platais, I., et al

    Meibom, S., Barnes, S. A., Platais, I., et al. 2015, Nature, 517, 589

    2015

  47. [55]

    D., Moya, A., Adibekyan, V ., et al

    Mena, E. D., Moya, A., Adibekyan, V ., et al. 2019, Astronomy & Astrophysics, 624, A78

    2019

  48. [56]

    S., Creevey, O

    Metcalfe, T. S., Creevey, O. L., & Christensen-Dalsgaard, J. 2009, The Astro- physical Journal, 699, 373

    2009

  49. [57]

    & Masuda, K

    Miyazaki, S. & Masuda, K. 2023, The Astronomical Journal, 166, 209

    2023

  50. [58]

    Moedas, N., Bossini, D., Deal, M., & Cunha, M. S. 2024, Astronomy & Astro- physics, 684, A113

    2024

  51. [59]

    Moedas, N., Nsamba, B., & Clara, M. T. 2020, in Dynamics of the Sun and Stars: Honoring the Life and Work of Michael J. Thompson, Springer, 259–265

    2020

  52. [60]

    2013, Astronomy & Astrophysics, 558, A106

    Mortier, A., Santos, N., Sousa, S., et al. 2013, Astronomy & Astrophysics, 558, A106

    2013

  53. [61]

    & López-Sastre, R

    Moya, A. & López-Sastre, R. J. 2022, Astronomy & Astrophysics, 663, A112

    2022

  54. [62]

    M., Delgado-Mena, E., et al

    Moya, A., Sarro, L. M., Delgado-Mena, E., et al. 2022, Astronomy & Astro- physics, 660, A15

    2022

  55. [63]

    J., & Davies, G

    Moya, A., Zuccarino, F., Chaplin, W. J., & Davies, G. R. 2018, The Astrophysical Journal Supplement Series, 237, 21

    2018

  56. [64]

    2016, Astronomy & Astrophysics, 593, A65

    Nissen, P. 2016, Astronomy & Astrophysics, 593, A65

    2016

  57. [65]

    Nissen, P. E. 2015, Astronomy & Astrophysics, 579, A52

    2015

  58. [66]

    1984, As- trophysical Journal, Part 1 (ISSN 0004-637X), vol

    Noyes, R., Hartmann, L., Baliunas, S., Duncan, D., & Vaughan, A. 1984, As- trophysical Journal, Part 1 (ISSN 0004-637X), vol. 279, April 15, 1984, p. 763-777. Research supported by the Smithsonian Institution and National Ge- ographic Society., 279, 763

    1984

  59. [67]

    2018, Monthly Notices of the Royal Astronomical Society, 477, 5052

    Nsamba, B., Campante, T., Monteiro, M., et al. 2018, Monthly Notices of the Royal Astronomical Society, 477, 5052

    2018

  60. [68]

    L., Monteiro, M

    Nsamba, B., Campante, T. L., Monteiro, M. J., Cunha, M. S., & Sousa, S. G. 2019, Frontiers in Astronomy and Space Sciences, 6, 25

    2019

  61. [69]

    2025, Monthly Notices of the Royal As- tronomical Society, 536, 2558

    Nsamba, B., Weiss, A., & Kamulali, J. 2025, Monthly Notices of the Royal As- tronomical Society, 536, 2558

    2025

  62. [70]

    B., et al

    Paxton, B., Schwab, J., Bauer, E. B., et al. 2018, The Astrophysical Journal Sup- plement Series, 234, 34

    2018

  63. [71]

    2019, The Astrophysical Journal Sup- plement Series, 243, 10

    Paxton, B., Smolec, R., Schwab, J., et al. 2019, The Astrophysical Journal Sup- plement Series, 243, 10

    2019

  64. [72]

    2011, the Journal of machine Learning research, 12, 2825

    Pedregosa, F., Varoquaux, G., Gramfort, A., et al. 2011, the Journal of machine Learning research, 12, 2825

    2011

  65. [73]

    2018, The exoplanet handbook (Cambridge university press)

    Perryman, M. 2018, The exoplanet handbook (Cambridge university press)

    2018

  66. [74]

    Pijpers, F. P. 2003, Astronomy & Astrophysics, 400, 241

    2003

  67. [75]

    H., An, D., Molenda-˙Zakowicz, J., et al

    Pinsonneault, M. H., An, D., Molenda-˙Zakowicz, J., et al. 2012, The Astrophys- ical Journal Supplement Series, 199, 30

    2012

  68. [76]

    H., Elsworth, Y ., Epstein, C., et al

    Pinsonneault, M. H., Elsworth, Y ., Epstein, C., et al. 2014, The Astrophysical Journal Supplement Series, 215, 19

    2014

  69. [77]

    1900, Entropie, 144, 164

    Planck, M. 1900, Entropie, 144, 164

    1900

  70. [78]

    V ., & Gulin, A

    Prokhorenkova, L., Gusev, G., V orobev, A., Dorogush, A. V ., & Gulin, A. 2018, Advances in neural information processing systems, 31

    2018

  71. [79]

    & Scaramella, R

    Pulone, L. & Scaramella, R. 1997, in Neural Nets WIRN VIETRI-96: Proceed- ings of the 8th Italian Workshop on Neural Nets, Vietri sul Mare, Salerno, Italy, 23–25 May 1996, Springer, 231–236 Ramírez, I., Melendez, J., & Asplund, M. 2009, Astronomy & Astrophysics, 508, L17

    1997

  72. [80]

    Rasio, F. A. & Ford, E. B. 1996, Science, 274, 954

    1996

  73. [81]

    M., Buldgen, G., Miglio, A., et al

    Rendle, B. M., Buldgen, G., Miglio, A., et al. 2019, Monthly Notices of the Royal Astronomical Society, 484, 771

    2019

  74. [82]

    & Nayfonov, A

    Rogers, F. & Nayfonov, A. 2002, The Astrophysical Journal, 576, 1064

    2002

  75. [83]

    M., & VandenBerg, D

    Sandage, A., Lubin, L. M., & VandenBerg, D. A. 2003, Publications of the As- tronomical Society of the Pacific, 115, 1187

    2003

  76. [84]

    2013, Astronomy & Astrophysics, 556, A150

    Santos, N., Sousa, S., Mortier, A., et al. 2013, Astronomy & Astrophysics, 556, A150

    2013

  77. [85]

    Schlafly, E. F. & Finkbeiner, D. P. 2011, The Astrophysical Journal, 737, 103

    2011

  78. [86]

    Scholz, F. W. & Stephens, M. A. 1987, Journal of the American Statistical As- sociation, 82, 918

    1987

  79. [87]

    & Mallen-Ornelas, G

    Seager, S. & Mallen-Ornelas, G. 2003, The Astrophysical Journal, 585, 1038

    2003

  80. [88]

    2017, The Astrophysical Journal Sup- plement Series, 233, 23 Silva Aguirre, V ., Ballot, J., Serenelli, A., & Weiss, A

    Serenelli, A., Johnson, J., Huber, D., et al. 2017, The Astrophysical Journal Sup- plement Series, 233, 23 Silva Aguirre, V ., Ballot, J., Serenelli, A., & Weiss, A. 2011, Astronomy and Astrophysics, 529, A63 Silva Aguirre, V ., Christensen-Dalsgaard, J., Chaplin, W., et al. 2...

    2017

  81. [89]

    2015, Nature, 517, 557

    Soderblom, D. 2015, Nature, 517, 557

    2015

  82. [90]

    Soderblom, D. R. 2010, Annual Review of Astronomy and Astrophysics, 48, 581 Student. 1908, Biometrika, 1

    2010

  83. [91]

    A., Bahcall, J

    Thoul, A. A., Bahcall, J. N., & Loeb, A. 1993, arXiv preprint astro-ph/9304005

    1993 arXiv

  84. [92]

    2019, Astronomy & Astrophysics, 622, A59

    Titarenko, A., Recio-Blanco, A., de Laverny, P., Hayden, M., & Guiglion, G. 2019, Astronomy & Astrophysics, 622, A59

    2019

  85. [93]

    2010, The Astronomy and Astrophysics Review, 18, 67

    Torres, G., Andersen, J., & Giménez, A. 2010, The Astronomy and Astrophysics Review, 18, 67

    2010

  86. [94]

    N., & Holman, M

    Torres, G., Winn, J. N., & Holman, M. J. 2008, The Astrophysical Journal, 677, 1324 Van Der Walt, S., Colbert, S. C., & Varoquaux, G. 2011, Computing in science & engineering, 13, 22

    2008

  87. [95]

    A., Brogaard, K., Leaman, R., & Casagrande, L

    VandenBerg, D. A., Brogaard, K., Leaman, R., & Casagrande, L. 2013, The As- trophysical Journal, 775, 134

    2013

  88. [96]

    2016, Monthly Notices of the Royal Astronomical Society, 461, 4206

    Verma, K., Hanasoge, S., Bhattacharya, J., Antia, H., & Krishnamurthi, G. 2016, Monthly Notices of the Royal Astronomical Society, 461, 4206

    2016

  89. [97]

    E., et al

    Virtanen, P., Gommers, R., Oliphant, T. E., et al. 2020, Nature methods, 17, 261

    2020

  90. [98]

    2006, in Robust Methods for Power System State Estimation and Load Forecasting

    Wehenkel, L., Ernst, D., & Geurts, P. 2006, in Robust Methods for Power System State Estimation and Load Forecasting

    2006

  91. [99]

    Wolpert, D. H. 1992, Neural networks, 5, 241 Article number, page 10 of 12 Kamulali et al.: Ages of giant-planet host stars Appendix A: Additional tests Appendix A.1: Comparison with other LEGACY results We conducted additional comparisons between the estimates from MAISTEP an...

    1992

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.