Pith. sign in

REVIEW 2 major objections 6 minor 2 cited by

The Eclipsing Binaries via Artificial Intelligence. II. Need for Speed in PHOEBE Forward Models

T0 review · 2 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A trained neural network surrogate for the PHOEBE binary-modeling code reproduces light curves with sub-percent error and runs MCMC sampling over 10,000 times faster.

desk verdict A solid, well-validated ANN surrogate for PHOEBE forward models within a restricted parameter box; the abstract overstates, but the core result is real and worth refereeing. read the letter →

arxiv 2412.11837 v2 pith:BHCMPQNG submitted 2024-12-16 astro-ph.SR astro-ph.EPastro-ph.GAcs.LG

Marcin Wrona , Andrej Prša This is my paper · ORCID
classification astro-ph.SRastro-ph.EPastro-ph.GAcs.LG
keywords BinarystarsEclipsingLightcurvesAstronomysoftwaredatamodelingNeuralnetworks
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

Fitting an eclipsing binary's light curve with a physical model is computationally expensive: an MCMC run that explores parameter posteriors can require millions of forward models and days of cluster time. This paper claims that a fully connected feedforward neural network, PHOEBAI, trained on more than a million PHOEBE-generated light curves, can replace the forward model itself for a six-parameter family of detached binaries, with systematic errors below about 1% and often as low as 0.01%. The measured payoff is concrete: on identical MCMC settings and hardware, the ANN finished in 2.16 seconds while PHOEBE took 21.63 hours, a speedup above four orders of magnitude. The paper also quantifies two limits of the approach: a network's intrinsic noise of roughly one part per thousand caps achievable parameter accuracy no matter how good the data are, and light contamination acts as a dilution that strongly biases every parameter unless it is fitted as a free variable. If this holds on real observations, the dominant bottleneck in eclipsing-binary analysis moves from computation to the design of complete, representative training sets.

What carries the argument

The load-bearing object is PHOEBAI, a fully connected feedforward ANN with six hidden layers of 512 nodes each, mapping six dimensionless eclipse-shape parameters ($e\sin\omega$, $e\cos\omega$, $\cos i$, $(R_1+R_2)/a$, $R_2/R_1$, $T_2/T_1$) to 501 normalized fluxes over orbital phases $-0.75$ to $0.25$. Once trained by regression on 1.25 million PHOEBE light curves (one million for training, 250,000 for validation), every forward evaluation is about $10^5$ cheap multiply-accumulate and activation operations—sub-millisecond—instead of PHOEBE's numerical surface discretization, atmosphere calls, and eclipse integration. The second piece of machinery is the dilution fraction $D = F_S/(F_S+F_B)$, applied after prediction as a flux-scaling parameter; including $D$ in the fit restores parameter recovery under light contamination.

What would settle it

Fit a set of detached eclipsing binaries with independently measured orbital and stellar parameters (from double-lined radial velocities and Gaia parallaxes) using PHOEBAI on their TESS light curves, and compare the recovered six parameters with the measured values; systematic mismatches above the claimed ~1% level, or accuracy collapsing for primaries outside the fixed F0 template, would falsify the surrogate's transfer claim.

Watch

Extended reading notes

Core claim

In the paper's own terms, a regression ANN can stand in for PHOEBE in the forward-model loop: feed it the six light-curve-shaping parameters—$e\sin\omega$, $e\cos\omega$, $\cos i$, $(R_1+R_2)/a$, $R_2/R_1$, and $T_2/T_1$—and it returns a 501-point phase-folded TESS light curve for a fixed F0 primary, with sub-percent systematic error across the training box. The chosen architecture, found by searching more than ten thousand configurations, has six hidden layers of 512 nodes with elu in the first hidden layer, sigmoid in the rest, a linear output, and the Adam optimizer. The claim is tested on synthetic data only: 100,000 unseen PHOEBE curves with realistic TESS noise are recovered with parameter accuracies around $10^{-3}$, the same behavior appears across five cross-validation networks, and the posterior distributions match PHOEBE's own sampling. The paper further claims that MCMC posterior scatter narrows as $N^{-1/2}$ with the number of data points while recovered parameter accuracy stays flat, so posterior width alone underestimates true uncertainty, and that adding a dilution fraction $D$ as a free parameter is necessary and sufficient to undo the bias that light contamination otherwise causes.

Load-bearing premise

The entire accuracy claim rests on PHOEBE's synthetic light curves being faithful stand-ins for real TESS observations of detached eclipsing binaries, so a network trained and tested only on PHOEBE output keeps its sub-percent errors when applied to nature.

Editorial extensions

If this is right

  • A single CPU core can now produce MCMC posteriors for one detached EB in minutes rather than days, so population-scale studies of TESS, Kepler, and OGLE eclipsing binaries become computationally feasible.
  • Because recovered parameter accuracy plateaus at the ANN's intrinsic noise near $10^{-3}$, collecting larger or higher-cadence light curves will not improve fitted parameters beyond that floor.
  • MCMC posterior widths shrink as $N^{-1/2}$ even when accuracy is unchanged, so reported parameter error bars from sampling alone systematically overstate precision.
  • Dilution must be included as a free parameter in AI-based EB fitting; without it, even weak contamination biases parameters substantially, with $R_2/R_1$ the most vulnerable.
  • The surrogate only interpolates: optimization or sampling that leaves the training box invalidates the solution, so training sets must be kept complete and representative of real EB parameter distributions.

Reading between the lines

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

  • The same six-layer architecture should transfer to other passbands and primary-star types by retraining on PHOEBE models; the limiting factor is the training distribution, not the network design.
  • A practical extension would be to train directly on diluted light curves rather than adding $D$ post hoc, which could reduce the $D$ degeneracies the paper reports for uncontaminated systems.
  • If the speedup survives contact with real data, catalog-scale reverse searches—fitting tens of thousands of systems to find rare configurations—become the natural next use of the surrogate.
  • The posterior-scatter result implies that published EB uncertainties from MCMC alone are systematically underestimated; adding the surrogate's systematic floor in quadrature would make error bars honest.
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

2 major / 6 minor

Summary. The paper presents PHOEBAI, a fully connected feedforward neural network trained on about 1.25 million PHOEBE-generated synthetic light curves to emulate the forward model for detached eclipsing binaries. The network maps six parameters (e sin ω, e cos ω, cos i, (R1+R2)/a, R2/R1, T2/T1) to a 501-point TESS-band light curve. The authors select a 6×512 architecture via a large hyperparameter search, then validate it with five-fold cross-validation, noise-injection studies spanning six decades of uncertainty, a held-out set of 100,000 systems, data-size scaling tests, and dilution studies with about 10,000 fits. They report a speedup of more than four orders of magnitude in MCMC sampling (2.16 s vs. 21.63 hr on the same hardware with identical sampler settings) and systematic parameter errors below 1%, often near 0.01%, within the training box. The paper also demonstrates that including a dilution fraction as a free parameter is essential for accurate parameter recovery when light contamination is present.

Significance. If the reported accuracy and speed hold, PHOEBAI is a practical drop-in surrogate for PHOEBE in optimization and MCMC sampling for the six-parameter detached-EB problem, enabling population-scale analyses that are currently prohibitively expensive. The paper's strengths are its extensive internal validation: five-fold cross-validation with five independently trained networks, noise injection over six decades, 100,000 unseen systems, data-size scaling, and roughly 10,000 dilution fits. The timing comparison is concrete and fair, using identical sampler settings and hardware. The paper also honestly discusses limitations, notably in Section 6, where it states that additional networks must be trained for different primary stars and passbands. The main caveat is that the accuracy claims are validated only on synthetic light curves drawn from the same filtered, fixed-primary, TESS training distribution, so the headline 'entire parameter space' claim needs to be scoped accordingly.

major comments (2)
  1. [Abstract and Section 5.1] The abstract's claim of 'systematic errors not exceeding 1% ... across the entire parameter space' is not supported as stated. The network is trained and tested with the primary fixed to M1=1.61 Msun, R1=1.728 Rsun, T1=7220 K (Section 5.1), a single passband (TESS T, Section 4.2), and a filtered parameter distribution that enforces R2<=R1, T2<=T1, eclipse amplitude >0.01, and no Roche-lobe overflow. The 'entire parameter space' is therefore the filtered, fixed-primary, TESS training box, not the space of real eclipsing binaries. The sentence in Section 5.1 that fixing the primary 'does not cause any loss of generality' is an unsupported assertion; Section 6 correctly acknowledges that new networks are needed for different primaries. Please revise the abstract and conclusions to state 'within the parameter space used for training' and remove or substantially qualify the 'no loss of generality' sentence.
  2. [Section 5.3.2, Table 3, and Section 6] The errors reported in Table 3 and described in Section 6 as 'systematic errors' are end-to-end parameter-recovery errors on synthetic data. They include the ANN approximation error, the optimizer/sampler's ability to find the best model, parameter degeneracies, and the effect of edge-of-range solutions. The paper never reports a direct comparison of ANN-predicted fluxes against PHOEBE fluxes on a held-out set, so the 'intrinsic noise' level of the ANN is inferred rather than measured. Please clarify in the text that Table 3 quantifies the total recovery floor, not the ANN forward-model error alone, and state whether direct flux-level metrics (e.g., MSE or MedAE on the held-out set) were computed; if they were, reporting them would strengthen the paper, and if not, the systematic-error language should be qualified accordingly.
minor comments (6)
  1. [Section 5.4] The dilution model is described verbally but no equation is given for how the ANN-predicted light curve is modified when D is a free parameter. A formula such as F_obs = D * F_ann + (1-D) * B, or the equivalent, would remove ambiguity about whether D multiplies the entire ANN output or acts on the variability amplitude.
  2. [Section 4.1] The rule of thumb '10 per degree of freedom' is said to imply a sample size of about one million for six parameters, but 10 per degree of freedom would give 60. This sentence is confusing and should be reworded or corrected, perhaps by clarifying that the one-million figure comes from the required density and nonlinearity of the mapping rather than this simple rule.
  3. [Table 3] Table 3 mixes absolute errors (e sin ω, e cos ω, cos i) and relative errors ((R1+R2)/a, R2/R1, T2/T1) without stating this in the caption. Please specify which columns are absolute and which are relative to avoid misinterpretation of the '1%' claim.
  4. [Section 5.3.1] The cross-validation description says 'five equal parts (validation sets)' but then states each training set has one million entries and each validation set has 250,000. Please clarify that the five validation sets are disjoint 250,000-entry subsets of the 1.25 million dataset, with the remaining one million used for training in each fold.
  5. [Figure 15] The lower panel shows fitted lines of the form A/sqrt(N) for the posterior scatter, but the fit parameters A and any uncertainties are not reported. Adding the fitted values would let readers reproduce the scaling claim.
  6. [References] The reference list contains a duplicate entry for Virtanen et al. 2020. Please remove the duplication.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: PHOEBAI is a standard, internally consistent emulator of PHOEBE validated on held-out synthetic light curves; the abstract's scope is narrower than its wording, but that is a generalization limitation, not circularity.

full rationale

The paper's central claim is that a fixed 6x512 feedforward ANN can reproduce PHOEBE forward models for detached eclipsing binaries in the six-parameter training box, with a >10^4 speedup and ~1e-3-level parameter recovery on synthetic data. This claim is not circular: the ANN is a supervised regression emulator, and its accuracy is tested on 100,000 held-out PHOEBE light curves never used in training, plus five-fold cross-validation (Secs. 5.3.1, 5.3.3). The 'systematic errors' are emulator approximation and inversion errors relative to known synthetic truth, not astrophysical validation; Table 3 and Sec. 6 characterize the intrinsic noise floor of ~1e-3 and state that new networks are needed for other primaries and passbands. The speedup is measured directly with identical samplers and hardware (Sec. 6). The paper explicitly discloses the fixed F0 primary (M1=1.61 Msun, R1=1.728 Rsun, T1=7220 K, Sec. 5.1) and the single TESS band (Sec. 4.2), so the abstract's 'entire parameter space' is the filtered training box; this is an overstatement of scope but does not make the derivation self-referential. The only self-citations (Prsa et al. 2008 for the parameter set, Prsa et al. 2022 for TESS noise) are contextual and not load-bearing: the surrogate's performance does not rest on those citations being true. No equation is defined in terms of the quantity it purports to predict, and no fitted parameter is renamed as a prediction. The emulator accuracy is genuinely evaluated on inputs outside the training set, even though those inputs share the same generator; that is standard surrogate validation, not circularity. Transfer to real observations is an open limitation, not a circular step.

Assumptions & free parameters 3 free parameters · 3 assumptions · 0 invented entities

The paper introduces no new physical entities. The load-bearing assumptions are the fidelity of PHOEBE, the representativeness of the hand-chosen training distributions, and the validity of the filtered training box. The ANN weights are fitted but are the model itself, not an independent free parameter of the argument.

free parameters (3)
  • Training-set parameter distributions = e ~ U(0,0.8); cos i ~ U(cos 90, cos 55); (R1+R2)/a ~ U(0.04,0.75); M2 ~ U(0.59,1.61) Msun; log-normal spreads 0.075…
    Chosen by hand from literature and filtered by eclipse criteria. The ANN's completeness boundary and representativeness, hence its systematic errors, are defined by these chosen ranges and densities (Sec. 4, 5.1).
  • Fixed primary stellar parameters = M1=1.61 Msun, R1=1.728 Rsun, T1=7220 K
    The primary is fixed to an F0 star for the entire training set. The paper asserts this causes no loss of generality because new networks can be trained, but all reported accuracies apply only to this primary and the TESS band (Sec. 5.1).
  • Low-noise threshold for reporting systematic errors = sigma <= 1e-3 in relative flux
    The paper defines the ANN's 'intrinsic noise' floor at about 1e-3 and reports Table 3 only for data with noise below this threshold. The 'systematic errors not exceeding 1%' claim is conditioned on this hand-chosen threshold (Sec. 5.3.2, Table 3).
assumptions (3)
  • domain assumption PHOEBE 2.4.16 is a faithful forward model for detached eclipsing binary light curves in the TESS band, including Roche geometry, Castelli & Kurucz atmospheres, gravity darkening, and reflection.
    Invoked in Sec. 5.1 to generate the training set. All ANN accuracies are measured against PHOEBE; if PHOEBE is systematically wrong, the ANN inherits the bias.
  • domain assumption The sampled parameter distributions (based on Raghavan et al. 2010, Duchene & Kraus 2013, Moe & Di Stefano 2017, Pecaut & Mamajek 2013) are representative of the real detached EB population, and interpolation within the filtered training box is sufficient.
    Stated in Sec. 4 as the completeness and representativeness criteria. The ANN is trained on this box and the authors require the optimizer and sampler to stay inside it (Sec. 5.2).
  • ad hoc to paper Exclusion filters (eclipse amplitude > 0.01, no Roche lobe overflow, R2 <= R1, T2 <= T1) define a physically valid but biased subset of the parameter space; the ANN is validated only on this filtered subset.
    Sec. 5.1 and Fig. 4 show the filtered distributions differ from the priors. The 'entire parameter space' claim in the abstract refers to the filtered box, not the full physical parameter space.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Eclipsing Binaries via Artificial Intelligence. II. Need for Speed in PHOEBE Forward Models." pith.science (2026). https://pith.science/paper/BHCMPQNG

@misc{pith2026241211837,
  author       = {Pith},
  title        = {Pith review of: The Eclipsing Binaries via Artificial Intelligence. II. Need for Speed in PHOEBE Forward Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BHCMPQNG}},
  note         = {Machine review of arXiv:2412.11837}
}
read the original abstract

In modern astronomy, the quantity of data collected has vastly exceeded the capacity for manual analysis, necessitating the use of advanced artificial intelligence (AI) techniques to assist scientists with the most labor-intensive tasks. AI can optimize simulation codes where computational bottlenecks arise from the time required to generate forward models. One such example is PHOEBE, a modeling code for eclipsing binaries (EBs), where simulating individual systems is feasible, but analyzing observables for extensive parameter combinations is highly time-consuming. To address this, we present a fully connected feedforward artificial neural network (ANN) trained on a dataset of over one million synthetic light curves generated with PHOEBE. Optimization of the ANN architecture yielded a model with six hidden layers, each containing 512 nodes, provides an optimized balance between accuracy and computational complexity. Extensive testing enabled us to establish ANN's applicability limits and to quantify the systematic and statistical errors associated with using such networks for EB analysis. Our findings demonstrate the critical role of dilution effects in parameter estimation for EBs, and we outline methods to incorporate these effects in AI-based models. This proposed ANN framework enables a speedup of over four orders of magnitude compared to traditional methods, with systematic errors not exceeding 1\%, and often as low as 0.01\%, across the entire parameter space.

Figures

Figures reproduced from arXiv: 2412.11837 by the authors.

Figure 1
Figure 1. A schematic of the feedforward neural network. The IL (leftmost collection of nodes p1, . . . , pn) takes EB parameters as input; these parameters are then propagated to the hidden layer (middle collection of nodes h1 . . . hm) and, in turn, to the OL (rightmost collection of nodes o1 . . . oq). The mapping is uniquely defined by the set of weights w11 . . . wnm, v11 . . . vmq, and biases b1 . . . bm, c1 . . . cq ad… view at source ↗
Figure 2
Figure 2. Overlay of light curves generated for a test sample. For each phase, the median and 18th–84th percentile range of fluxes were computed. The red solid line represents the median light curve, while the shaded region between the red dashed lines indicates the percentile range. Each thin black line represents an individual light curve. Although only 5000 out of the 1,250,000 generated light curves are shown here, the st… view at source ↗
Figure 3
Figure 3. depicts a single RSCV run, illustrating the performance of various AFs in the output layer (OL). The figure demonstrates that the linear AF accounts for the majority of favorable results (with lower values indicating better performance). We performed similar analyses across multiple ANN architectures; by archi￾tecture, we refer to both the structure of the ANN, en￾compassing the number of hidden layers (HLs) and the… view at source ↗
Figures from the paper (16 more)
Figure 4
Figure 4. Figure 4: Density histograms comparing the initial and filtered distributions of the main set of parameters during the step of generating forward models using PHOEBE. was fixed at M1 = 1.61 M⊙, radius at R1 = 1.728 R⊙, and effective temperature at T1 = 7220 K, correspond￾ing to …
Figure 5
Figure 5. Figure 5: (Top) Phase-folded light curve of one synthetic dataset with fitted models. The dashed red line represents the model obtained using an ANN, while the solid blue line represents the model obtained with PHOEBE. (Bottom) Residuals after subtracting the ANN model fluxes fr…
Figure 6
Figure 6. Figure 6: Histograms of the differences between the ob￾tained parameter values and the true values derived using the bootstrapping method with PHOEBE. Black dashed lines denote the zero-point. This step highlights a major advantage of using ANNs over PHOEBE for model generation.…
Figure 7
Figure 7. Figure 7: Corner plot of combined posterior distributions, comparing results from the bootstrap analysis for the final solution using ANNs (red) and PHOEBE (blue). Standard deviations are provided above each histogram. Black dots indicate the true parameter values used to genera…
Figure 8
Figure 8. Figure 8: Architecture of the best-performing deep ANN based on cross-validation tests, using 1,000,000 objects for the training set and 250,000 for the validation set. This ANN contains six hidden layers (HLs), each with 512 nodes. The activation function (AF) for the first HL …
Figure 9
Figure 9. Figure 9: (Top) Learning curves for an ANN trained on five different subsets of the training set as part of a cross￾validation test. In each variant, the training set consisted of one million entries, and the validation set of 250,000. In each variant, the validation set consist…
Figure 10
Figure 10. Figure 10: Phase-folded synthetic light curves that were used to evaluate the stability and performance of ANNs dur￾ing cross-validation. 5.3.3. The impact of parameter values During model generation for a new dataset, we kept 100,000 objects separate from the training sessions.…
Figure 11
Figure 11. Figure 11: (Top) Corner plot with posterior distributions after sampling the solutions found by all five ANNs during the cross-validation test (colors match those used for learning curves in [PITH_FULL_IMAGE:figures/full_fig_p015_11.png]
Figure 12
Figure 12. Figure 12: The relationship between the injected noise level (Gaussian, with the x-axis representing the noise σ) and the parameter shifts and scatter for one synthetic system. (Top) Parameter shift, defined as the absolute or relative difference between the true parameter value…
Figure 13
Figure 13. Figure 13: 2D histograms showing the differences between the true parameter values and the highest log-probability estimates. The 2D histograms represent the average parameter shifts and scatter obtained using all five ANNs trained during the cross￾validation test. The top row d…
Figure 14
Figure 14. Figure 14: Phase-folded light curves of one synthetic object plotted alongside their respective PHOEBAI-fitted models. The left-hand plot shows the light curve with 100 data points, while the right-hand plot shows the same curve with 10,000 data points. Despite the large differe…
Figure 15
Figure 15. Figure 15: (Top) Median and min–max range of absolute and relative shifts of the parameters obtained using PHOE￾BAI for a group of 10 synthetic light curves, plotted against the sample size. For each sample size, all shifts are aver￾aged over the five ANNs trained during the cro…
Figure 17
Figure 17. Figure 17: Median parameter shifts as a function of the dilution fraction. Red/green lines represent cases with￾out/with D as a free parameter in the model. For each D, we calculated the median and 16th–84th percentile range of the shift based on a sample of about 6000 synthetic…
Figure 18
Figure 18. Figure 18: Comparison of true vs. recovered parameters after modeling about 6000 synthetic light curves using PHOEBAI. Green/red points represent results with/without the dilution factor as a free parameter. Black dashed lines show the y = x line. The upper/lower sets of plots s…
Figure 19
Figure 19. Figure 19: Shifts in the obtained parameters (e, ω, i regime) relative to the dilution factor. All points represent modeling processes that included D as a free parameter. For each D, we calculated the median and 16th–84th percentile range of shifts based on a sample of 10,000 s…
Figure 21
Figure 21. Figure 21: Comparison of true vs. predicted values of the dilution factor based on modeling 10,000 synthetic light curves. Red points denote the median value of the obtained D for a given true value, while error bars represent the 5th and 95th percentiles of this distribution. T…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. The present and the future of modeling eclipsing binary systems

    astro-ph.SR 2025-01 accept novelty 4.0 of 10

    A review of eclipsing binary modeling argues for separate boutique and bulk approaches and demonstrates a neural-network emulator that runs much faster than the physical model.

  2. From stellar light to astrophysical insight: automating variable star research with machine learning

    astro-ph.IM 2025-07 unverdicted

    An invited review of machine learning for automated variable star research, covering data cleaning, variability classification, stellar parameter inference, and foundation models.

Reference graph

Works this paper leans on

50 extracted references · 13 canonical work pages · cited by 2 Pith papers

  1. [1]

    Intrinsic Colors, Temperatures, and Bolometric Corrections of Pre-main-sequence Stars

    thebibliography [1] 20pt to REFERENCES 6pt =0pt -12pt 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 E...

    arXiv

  2. [2]

    2015, TensorFlow : Large-Scale Machine Learning on Heterogeneous Systems

    Abadi, M., Agarwal, A., Barham, P., et al. 2015, TensorFlow : Large-Scale Machine Learning on Heterogeneous Systems. https://www.tensorflow.org/

    2015

  3. [3]

    2011, in American Institute of Physics Conference Series, Vol

    Arenou , F. 2011, in American Institute of Physics Conference Series, Vol. 1346, American Institute of Physics Conference Series, ed. J. A. Docobo , V. S. Tamazian , & Y. Y. Balega , 107--121, 10.1063/1.3597593

    work page doi:10.1063/1.3597593 2011

  4. [4]

    M., Sip o cz , B

    Astropy Collaboration , Price-Whelan , A. M., Sip o cz , B. M., et al. 2018, , 156, 123, 10.3847/1538-3881/aabc4f

    doi:10.3847/1538-3881/aabc4f 2018

  5. [5]

    Bailer-Jones , C. A. L., Rybizki , J., Fouesneau , M., Mantelet , G., & Andrae , R. 2018, , 156, 58, 10.3847/1538-3881/aacb21

    doi:10.3847/1538-3881/aacb21 2018

  6. [6]

    2019, Canadian Association of Radiologists Journal, 70, 344, 10.1016/j.carj.2019.06.002

    Balki, I., Amirabadi, A., Levman, J., et al. 2019, Canadian Association of Radiologists Journal, 70, 344, 10.1016/j.carj.2019.06.002

    doi:10.1016/j.carj.2019.06.002 2019

  7. [7]

    F., et al

    Barclay , T., Huber , D., Rowe , J. F., et al. 2012, , 761, 53, 10.1088/0004-637X/761/1/53

    doi:10.1088/0004-637x/761/1/53 2012

  8. [8]

    J., Koch , D., Basri , G., et al

    Borucki , W. J., Koch , D., Basri , G., et al. 2010, Science, 327, 977, 10.1126/science.1185402

    doi:10.1126/science.1185402 2010

Show all 50 references
  1. [9]

    Castelli , F., & Kurucz , R. L. 2004, ArXiv Astrophysics e-prints

    2004

  2. [10]

    2015, Keras, https://keras.io

    Chollet, F., et al. 2015, Keras, https://keras.io

    2015

  3. [11]

    2011, , 529, A75, 10.1051/0004-6361/201116451

    Claret , A., & Bloemen , S. 2011, , 529, A75, 10.1051/0004-6361/201116451

    2011 doi

  4. [12]

    E., Pr s a , A., Stassun , K

    Conroy , K. E., Pr s a , A., Stassun , K. G., et al. 2014, , 147, 45, 10.1088/0004-6256/147/2/45

    2014 doi

  5. [13]

    E., Kochoska , A., Hey , D., et al

    Conroy , K. E., Kochoska , A., Hey , D., et al. 2020, , 250, 34, 10.3847/1538-4365/abb4e2

    2020 doi

  6. [14]

    2023, Gaia Data Release Scenario, https://www.cosmos.esa.int/web/gaia/release

    Consortium, G. 2023, Gaia Data Release Scenario, https://www.cosmos.esa.int/web/gaia/release

    2023

  7. [15]

    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

  8. [16]

    2017, , 154, 220, 10.3847/1538-3881/aa9332

    Foreman-Mackey , D., Agol , E., Ambikasaran , S., & Angus , R. 2017, , 154, 220, 10.3847/1538-3881/aa9332

    2017 doi

  9. [17]

    W., Lang , D., & Goodman , J

    Foreman-Mackey , D., Hogg , D. W., Lang , D., & Goodman , J. 2013, , 125, 306, 10.1086/670067

    2013 doi

  10. [18]

    A., & Skapura, D

    Freeman, J. A., & Skapura, D. M. 1991, Neural networks: algorithms, applications, and programming techniques (Redwood City, CA, USA: Addison Wesley Longman Publishing Co., Inc.)

    1991

  11. [19]

    Gaia Collaboration , Brown , A. G. A., Vallenari , A., et al. 2021, , 649, A1, 10.1051/0004-6361/202039657

    2021 doi

  12. [20]

    2010, Communications in Applied Mathematics and Computational Science, 5, 65, 10.2140/camcos.2010.5.65

    Goodman , J., & Weare , J. 2010, Communications in Applied Mathematics and Computational Science, 5, 65, 10.2140/camcos.2010.5.65

    2010 doi

  13. [21]

    M., Kurtz , D

    Hambleton , K. M., Kurtz , D. W., Pr s a , A., et al. 2013, , 434, 925, 10.1093/mnras/stt886

    2013 doi

  14. [22]

    2015, in Astrophysics and Space Science Library, Vol

    Huber , D. 2015, in Astrophysics and Space Science Library, Vol. 408, Giants of Eclipse: The Aurigae Stars and Other Binary Systems, 169, 10.1007/978-3-319-09198-3_7

    2015 doi

  15. [23]

    Hunter , J. D. 2007, Computing in Science and Engineering, 9, 90, 10.1109/MCSE.2007.55

    2007 doi

  16. [24]

    Lucy , L. B. 1967, , 65, 89

    1967

  17. [25]

    Luri , X., Brown , A. G. A., Sarro , L. M., et al. 2018, , 616, A9, 10.1051/0004-6361/201832964

    2018 doi

  18. [26]

    2017, , 230, 15, 10.3847/1538-4365/aa6fb6

    Moe , M., & Di Stefano , R. 2017, , 230, 15, 10.3847/1538-4365/aa6fb6

    2017 doi

  19. [27]

    Oliphant, T. E. 2006, A guide to NumPy, Vol. 1 (Trelgol Publishing USA)

    2006

  20. [28]

    Orosz , J. A. 2015, in Astronomical Society of the Pacific Conference Series, Vol. 496, Living Together: Planets, Host Stars and Binaries, ed. S. M. Rucinski , G. Torres , & M. Zejda , 55. 1503.07295

    2015 arXiv

  21. [29]

    J., & Mamajek , E

    Pecaut , M. J., & Mamajek , E. E. 2013, , 208, 9, 10.1088/0067-0049/208/1/9

    2013 doi

  22. [30]

    2011, Journal of Machine Learning Research, 12, 2825

    Pedregosa, F., Varoquaux, G., Gramfort, A., et al. 2011, Journal of Machine Learning Research, 12, 2825

    2011

  23. [31]

    F., Devinney , E

    Pr s a , A., Guinan , E. F., Devinney , E. J., et al. 2008, , 687, 542, 10.1086/591783

    2008 doi

  24. [32]

    2005, , 628, 426, 10.1086/430591

    Pr s a , A., & Zwitter , T. 2005, , 628, 426, 10.1086/430591

    2005 doi

  25. [33]

    2018, Modeling and Analysis of Eclipsing Binary Stars; The theory and design principles of PHOEBE (Institute of Physics), 10.1088/978-0-7503-1287-5

    Pr s a , A. 2018, Modeling and Analysis of Eclipsing Binary Stars; The theory and design principles of PHOEBE (Institute of Physics), 10.1088/978-0-7503-1287-5

    2018 doi

  26. [34]

    2019, , 131, 068001, 10.1088/1538-3873/ab0f41

    Pr s a , A., Zhang , M., & Wells , M. 2019, , 131, 068001, 10.1088/1538-3873/ab0f41

    2019 doi

  27. [35]

    E., et al

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

    2022 doi

  28. [36]

    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

  29. [37]

    R., Winn , J

    Ricker , G. R., Winn , J. N., Vanderspek , R., et al. 2015, Journal of Astronomical Telescopes, Instruments, and Systems, 1, 014003, 10.1117/1.JATIS.1.1.014003

    2015 doi

  30. [38]

    2021, arXiv e-prints, arXiv:2111.15006, 10.48550/arXiv.2111.15006

    Ridden-Harper , R., Rest , A., Hounsell , R., et al. 2021, arXiv e-prints, arXiv:2111.15006, 10.48550/arXiv.2111.15006

    2021 doi arXiv

  31. [39]

    2022, in Artificial Intelligence, ed

    Roy, S., & Dubey, J. 2022, in Artificial Intelligence, ed. A. A. Sk, T. Turki, T. K. Ghosh, S. Joardar, & S. Barman (Cham: Springer Nature Switzerland), 191--201

    2022

  32. [40]

    Russell , H. N. 1948, in Harvard Observatory Monographs, Vol. 7 (Harvard Observatory), 181

    1948

  33. [41]

    W., Pr s a , A., Welsh , W

    Slawson , R. W., Pr s a , A., Welsh , W. F., et al. 2011, , 142, 160, 10.1088/0004-6256/142/5/160

    2011 doi

  34. [42]

    2014, , 147, 86, 10.1088/0004-6256/147/4/86

    Tokovinin , A. 2014, , 147, 86, 10.1088/0004-6256/147/4/86

    2014 doi

  35. [43]

    2010, , 18, 67, 10.1007/s00159-009-0025-1

    Torres , G., Andersen , J., & Gim \'e nez , A. 2010, , 18, 67, 10.1007/s00159-009-0025-1

    2010 doi

  36. [45]

    E., et al

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

    2020 doi

  37. [46]

    Vogt , F. P. A., Bond , H. E., Cracraft , M. M., et al. 2014, in American Astronomical Society Meeting Abstracts, Vol. 223, American Astronomical Society Meeting Abstracts \#223, 154.18

    2014

  38. [47]

    2017, , 129, 065003, 10.1088/1538-3873/aa682a

    Wells , M., Pr s a , A., Jones , L., & Yoachim , P. 2017, , 129, 065003, 10.1088/1538-3873/aa682a

    2017 doi

  39. [48]

    A., & Pr s a , A

    Wells , M. A., & Pr s a , A. 2021, , 253, 32, 10.3847/1538-4365/abd5ba

    2021 doi

  40. [49]

    F., Orosz , J

    Welsh , W. F., Orosz , J. A., Short , D. R., et al. 2015, , 809, 26, 10.1088/0004-637X/809/1/26

    2015 doi

  41. [50]

    Wilson , R. E. 1990, , 356, 613, 10.1086/168867

    1990 doi

  42. [51]

    E., & Devinney , E

    Wilson , R. E., & Devinney , E. J. 1971, , 166, 605, 10.1086/150986

    1971 doi

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.