Pith. sign in

REVIEW 4 major objections 5 minor 22 references

SPINN: An Optimal Self-Supervised Physics-Informed Neural Network Framework

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The paper claims that a self-supervised physics-informed neural network with a learned physics-coefficient neuron estimates liquid-sodium Nusselt numbers in miniature heat sinks strictly within an 8% error using only 87 CFD data points.

desk verdict Useful benchmark of ML surrogates for Na heat-sink Nusselt numbers, but the self-supervised PINN claim is unsupported by an ill-posed loss and by the paper's own numbers. read the letter →

arxiv 2509.05886 v1 pith:KYI6Y6XV submitted 2025-09-07 cs.LG

Reza Pirayeshshirazinezhad This is my paper · ORCID
classification cs.LG
keywords physics-informedneuralnetworkself-supervisedlearningNusseltnumberliquidsodiumminiatureheatsinktransfersurrogatemodeluncertaintyquantification
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 aims to show that a small, physics-informed neural network can replace days-long CFD simulations for predicting how well liquid sodium transfers heat in miniature rectangular heat sinks. Using only 87 CFD-generated data points, it trains kernel methods, ordinary neural networks, a transfer-learned network, and a self-supervised physics-informed network, then compares their holdout errors. The central assertion is that the self-supervised PINN, whose loss couples a data-fidelity term with a physics-correlation term through a learned physics-coefficient neuron, stays strictly within an 8% error margin and is more robust to random weight initialization than ordinary neural networks. The paper also claims that transferring the first layer of a water-trained network to the sodium network slightly improves accuracy, and that all methods are mostly within 8% for this dataset. If true, the framework gives engineers a fast surrogate for designing liquid-metal-cooled heat sinks where experiments are scarce and CFD costs about three days per point.

What carries the argument

The load-bearing object is the self-supervised PINN architecture: a fully connected network with ReLU hidden units, a linear output for $Nu_{ave}$, a sigmoid physics-coefficient neuron, and a two-term loss. The first term is MAPE between predicted and CFD Nusselt numbers; the second is MSE between the coefficient and the value of an empirical heat-transfer correlation $\hat{Nu}_{ave}$ (Eq. 11) evaluated from the inputs. The sigmoid keeps the coefficient in (0,1), and Bayesian optimization tunes the layer sizes and learning rate. The intended mechanism is that the coefficient learns how much to trust the physics correlation versus the data, so the network is self-supervised by the physics formula rather than by labeled physics outputs. Supporting machinery includes 10-fold cross-validation for hyperparameter selection, 500-run Monte Carlo simulation to measure variance of predictions and MAPE, and genetic-algorithm selection of which layers to transfer from a water-trained network.

What would settle it

A reader could settle the central claim by ablating the physics term: train the same architecture with only the MAPE loss and compare holdout error and variance; if removing the MSE physics term leaves the 8% bound and the low variance unchanged, the physics coefficient is not doing the work attributed to it. In addition, replacing Eq. (11) with an independent correlation and re-checking the 8% bound would test whether the physics term is an independent constraint.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that a self-supervised PINN with a dedicated physics-coefficient neuron can estimate the average Nusselt number ($Nu_{ave}$) of liquid sodium in SS-316 rectangular miniature heat sinks within a strict 8% error on a holdout dataset, using a training set of just 87 CFD simulations. The network outputs both the Nusselt number and a scalar coefficient bounded between 0 and 1; the loss is MAPE between the output and the data plus MSE between the physics-based correlation $\hat{Nu}_{ave}$ (Eq. 11) and that coefficient. The paper interprets this as balancing data and physics according to their uncertainty, and reports that the PINN has about half the prediction variance of ordinary NN (max variance 0.1146 vs 0.2323) while remaining within the 8% band. A second claimed result is that transfer learning from a water-trained shallow NN, specifically the first layer, reduces MAPE from 0.0028 to 0.0020 for sodium, confirming that the most general layer transfers best. The paper positions the combined data-physics approach as more robust than physics-only regression, which has 5% to 10% error, and calls the framework optimal for data-scarce heat transfer modeling.

Load-bearing premise

The paper assumes that the scalar physics-coefficient neuron actually couples the physics correlation into the Nusselt-number prediction, but the loss only compares that coefficient to the correlation's value and never states how the coefficient changes the output; the correlation was itself fitted to the same CFD data, so if that coupling is absent the physics-informed mechanism is empty.

Editorial extensions

If this is right

  • If the 8% holdout bound holds beyond this dataset, engineers can screen Na-cooled miniature heat sink designs with a neural surrogate instead of running multi-day CFD simulations.
  • The PINN's roughly halved prediction variance (0.1146 vs 0.2323 max variance) implies the physics term stabilizes the network against weight-initialization randomness, which matters when the model must be retrained for new geometries.
  • Transfer learning from water to sodium means a source-domain NN trained on an abundant fluid can jump-start a target-domain model for a scarce-data liquid metal, cutting training epochs and improving MAPE.
  • Since physics-only regression gives 5% to 10% error while the hybrid PINN goes under 8%, the results support ensemble-like combination of empirical correlations with ML rather than choosing one over the other.

Reading between the lines

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

  • A testable extension is to make the physics coefficient actually gate the output, for example $Nu_{pred} = c \cdot \hat{Nu}_{ave} + (1-c) \cdot Nu_{NN}$; the current loss only compares the coefficient to the correlation, so the claimed physics-informed improvement may come from regularization rather than a hard physical constraint.
  • Because Eq. (11) was fitted to the same 87-point CFD dataset, the physics term is not an independent source of information; re-testing the method with a correlation derived from separate experiments or from first-principles dimensionless analysis would clarify whether the physics term adds real constraint.
  • The strict-within-8% claim rests on a single holdout from one CFD campaign; a natural stress test is to hold out entire flow regimes, such as all turbulent cases, and see whether the PINN extrapolates, or to apply the same architecture to another liquid metal such as NaK or lead-bismuth eutectic.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper develops machine-learning surrogate models for predicting the average Nusselt number of liquid sodium in miniature heat sinks from a small dataset of 87 CFD simulations. It compares Gaussian process, support vector regression, a shallow neural network, a transfer-learning variant, and a proposed 'self-supervised' physics-informed neural network (PINN) with an added physics-coefficient neuron. The central claims are that this SPINN framework is 'optimal', that it balances data and physics based on their uncertainty, that it achieves errors strictly within 8% on a holdout set, and that it is more robust than other ML methods.

Significance. If the method worked as described, the paper would offer a pragmatic surrogate for an expensive CFD problem and a possible template for combining empirical correlations with neural networks on small datasets. The paper deserves credit for a structured validation protocol: 10-fold cross-validation, Monte Carlo simulation with 500 evaluations, and systematic comparison across several algorithms. The application to liquid-sodium heat transfer is practically motivated. However, the central physics-informed mechanism is not correctly implemented, the 'physics' term is an empirical fit to the same dataset rather than an independent constraint, and the paper's own benchmark table contradicts the 'optimal' designation. The significance of the contribution as it stands is therefore low.

major comments (4)
  1. [Sec. 4.3, Eq. (12)] The loss function in Eq. (12) is ill-posed: it adds MSE(physics prediction, physics coefficient), where the physics prediction Nu_hat_ave from Eq. (11) lies between roughly 2.2 and 3.85 (Table 1), while the physics coefficient is a sigmoid neuron bounded to [0,1] by Eq. (14). This MSE term cannot be minimized to zero, and its gradient does not pull the Nusselt prediction toward the physics correlation. The claimed balancing of data and physics is therefore not implemented by the stated loss.
  2. [Sec. 2(e), Eqs. (8)-(11)] The 'physics' used in the PINN is the empirical correlation of Pourghasemi and Fathi [12], which was fitted to the same or closely related CFD data that constitutes the training set. Using this correlation as a self-supervised constraint on a holdout from that same 87-point dataset is circular: the physics term is not an independent physical law. The abstract and conclusion claim the model 'adheres to physical laws', but Eqs. (8)-(11) are a curve fit, not governing equations such as (1)-(4).
  3. [Table 5 and Sec. 4.3] The paper's central 'optimal' claim is internally contradicted by its own results: the 10-fold cross-validated MAPE of SPINN is 0.0185, about 6.6 times worse than the plain NN (0.0028) and worse than SVR-Bayesian (0.0125). Furthermore, the conclusion states that 'only self-supervised PINN remains within the 8% range', but the same section says all methods provide estimations 'mostly within an 8% margin of error', and Figure 11 shows all three NN-based methods inside the band. These statements are inconsistent.
  4. [Sec. 2(e) and Fig. 3] The manuscript says the physics-coefficient layer 'determines the weight of physics in the loss function to balance data and physics based on their uncertainty', but Eq. (12) is an unweighted sum of MAPE and MSE, and the coefficient is the target of the MSE, not a weight. The mechanism by which a scalar in [0,1] modulates the data-versus-physics trade-off, or is updated during backpropagation, is never specified. It is also unclear how the physics prediction Nu_hat_ave is computed for each input sample, since Eq. (8) appears garbled in the text.
minor comments (5)
  1. [Throughout] There are many typographical errors and unclear phrases, including 'Besian' instead of 'Bayesian', 'neroun' for 'neuron', 'MAPPE' for 'MAPE', 'the weight the of physics', and a garbled Eq. (8). The paper needs careful proofreading before any further consideration.
  2. [Sec. 2.1(c)] The text says the inputs are in R6, but only five inputs are listed (Pe, Dh, alpha, L, W). This inconsistency should be corrected or clarified.
  3. [Sec. 4.2] The text states 'Figure 3 illustrates that transferring the first layer results in the lowest MAPE error', but the relevant figure appears to be Figure 6; the cross-reference is wrong.
  4. [Sec. 2, Data preprocessing] The size of the holdout dataset is never stated. The validation section refers to a holdout set, but the number of points and how it was carved from the 87-point dataset are not described, which hampers reproducibility.
  5. [Sec. 4.3] The statement that SPINN achieves an error 'strictly within 8%' is ambiguous because MAPE is reported as 0.0185 (1.85%), while Figures 10-11 use an 8% band around the parity line. The relation between these two error metrics should be made explicit.

Circularity Check

2 steps flagged · score 6.0 of 10

The 'physics' in SPINN is an empirical correlation fitted in prior same-group work to the same 87 CFD data points, so the claimed physics-informed gain reduces to combining the data with a fit to the same data.

  1. fitted input called prediction [Section 2.1(e), Eqs. (11)-(12); Section 4.3]
    "The physics prediction of Nusselt number Nû ave [12, 17] is given as the following. ... 𝑁𝑢̂ 𝑎𝑣𝑒 = 𝑁𝑢∗(1.0 + 0.135(𝑃𝑒∗)0.388) (11) The loss function of self-supervised PINN is given as Loss = MAPE(prediction, labels) + MSE(physics prediction, physics coefficient) (12)."

    Eq. (11) is not an independent physical law: it is the analytical multivariate regression from [12], by the same group, fitted to the same CFD dataset used here ('Computational data [12] ... are used in this work to train ML models'). The PINN loss adds this fitted correlation as the 'physics prediction'. Therefore the physics constraint is a curve fit to the target Nuave labels, so the claimed improvement from 'physics' is partially forced by construction: the network is trained on the labels and simultaneously pulled toward a regression of the same labels. The paper's own statement that the combination 'is similar to ensemble learning' confirms the physics term is a second fitted model, not an independent benchmark.

  2. self citation load bearing [Section 2.1(e), Eq. (11); References [12], [17]; Acknowledgment]
    "The physics prediction of Nusselt number Nû ave [12, 17] is given as the following. ... The author thanks Prof. Nima Fathi and Prof. Mahyar Pourghasemi for valuable guidance and discussions during the stages of this project. Parts of this work are part of a broader collaborative effort, and a revised version including co-authors is in preparation for journal submission."

    The 'physics' in the central PINN contribution is sourced exclusively from [12] and [17], both authored by Pourghasemi and Fathi, the same researchers acknowledged as collaborators on this project. No external derivation, independent data source, or machine-checked verification of Eq. (11) is provided; the cited works supply both the CFD dataset and the correlation fitted to it. The paper's claim that SPINN is an 'optimal' physics-informed framework therefore rests on a self-citation chain whose cited content is itself the fitted result, not an external mathematical or experimental constraint.

full rationale

The central physics term, Eq. (11), is imported from same-group prior work [12,17] and is an empirical multivariate regression fitted to the same 87 CFD Nusselt numbers used for training and holdout evaluation. Using that fit as the 'physics prediction' in Eq. (12) means the purported physics constraint is a second model fitted to the target data, so the claimed robustness/accuracy gain is partly circular: it is an ensemble of the data with a fit of the same data. The acknowledgment identifies Fathi and Pourghasemi as collaborators, making the citation chain load-bearing rather than external. Independently of circularity, Eq. (12) as written is also ill-posed: the sigmoid-bounded physics coefficient lies in [0,1], while Table 1 shows Nuave ranging from about 2.5 to 3.85, so the MSE term compares incompatible scales and cannot implement the claimed balance. Internal contradictions (PINN 10-fold MAPE 0.0185 vs NN 0.0028 in Table 5; 'only self-supervised PINN remains within 8%' vs 'all ML methods provide fairly accurate estimations') further weaken the 'optimal' claim, but these are correctness issues rather than circularity. Because the central claim reduces in part to using a fitted correlation as 'physics,' the circularity score is 6, not higher, since the transfer-learning comparison and kernel benchmark results are self-contained and externally meaningful.

Assumptions & free parameters 4 free parameters · 4 assumptions · 1 invented entities

The central claim that SPINN is optimal rests on a learned physics coefficient that is fit during training, a physics correlation from the same group's prior work, and a validation set drawn from the same 87-point CFD dataset. The free parameters are the learned scalar, the architecture choices, and the reused empirical correlation constants. The assumptions are largely domain assumptions inherited from [12] and [14], plus the ambiguous coupling in Eq. (12).

free parameters (4)
  • Physics coefficient neuron (PC) = Approximately 0.5 (Figure 9)
    A learned scalar in the PINN loss that is supposed to balance data and physics; its exact role is ambiguous and it is fit during training, and the paper's own CV shows it does not improve over plain NN.
  • PINN architecture hyperparameters (neurons per layer, learning rate, optimizer) = 20-20-12 neurons, learning rate 0.34, Adam
    Selected by Bayesian optimization on the 87-point 10-fold CV, so the reported MAPE is an optimistic in-sample evaluation of the architecture.
  • Transfer learning architecture (layers transferred, neurons) = 2 hidden layers; first layer (3 neurons) transferred from water; second layer 8 random neurons
    Selected by GA on the Na dataset; the improvement over no transfer is small (0.0020 vs 0.0028).
  • Physics correlation constants in Eqs. (10)-(11) = 0.164, 10.2, -12, 0.135, 0.388
    Fitted to CFD data in prior work [12] by the same group; the present paper treats them as physics, but they are empirical regression coefficients for the same data regime.
assumptions (4)
  • domain assumption Equations (8)-(11) accurately represent the physics of Na forced convection in miniature heat sinks
    The correlation is taken from [12], an empirical fit to CFD data, not a first-principles law; using it as the physics target assumes it holds for all 87 points and beyond the fitted range.
  • domain assumption The 87 CFD data points from [12] are accurate ground truth
    No experimental validation is reported; the paper relies on the prior CFD framework of [12,17].
  • ad hoc to paper The water NN from [14] is a valid source for transfer learning to Na
    The water model was trained on nanofluid flow in a microchannel heat sink, which has different geometry, fluid, and input features; the paper assumes the first-layer features generalize.
  • standard math Standard incompressible Navier-Stokes and energy equations govern the CFD simulations
    Equations (1)-(4) are standard; no derivation issue.
invented entities (1)
  • Physics coefficient neuron (PC)
    purpose: A dedicated neuron that outputs a scalar in [0,1] to weight the physics contribution in the loss, intended to balance data and physics based on uncertainty
    No independent evidence that this mechanism improves performance; the paper's own CV shows plain NN outperforms it, and the loss definition is ambiguous.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SPINN: An Optimal Self-Supervised Physics-Informed Neural Network Framework." pith.science (2026). https://pith.science/paper/KYI6Y6XV

@misc{pith2026250905886,
  author       = {Pith},
  title        = {Pith review of: SPINN: An Optimal Self-Supervised Physics-Informed Neural Network Framework},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KYI6Y6XV}},
  note         = {Machine review of arXiv:2509.05886}
}
read the original abstract

A surrogate model is developed to predict the convective heat transfer coefficient of liquid sodium (Na) flow within rectangular miniature heat sinks. Initially, kernel-based machine learning techniques and shallow neural network are applied to a dataset with 87 Nusselt numbers for liquid sodium in rectangular miniature heat sinks. Subsequently, a self-supervised physics-informed neural network and transfer learning approach are used to increase the estimation performance. In the self-supervised physics-informed neural network, an additional layer determines the weight the of physics in the loss function to balance data and physics based on their uncertainty for a better estimation. For transfer learning, a shallow neural network trained on water is adapted for use with Na. Validation results show that the self-supervised physics-informed neural network successfully estimate the heat transfer rates of Na with an error margin of approximately +8%. Using only physics for regression, the error remains between 5% to 10%. Other machine learning methods specify the prediction mostly within +8%. High-fidelity modeling of turbulent forced convection of liquid metals using computational fluid dynamics (CFD) is both time-consuming and computationally expensive. Therefore, machine learning based models offer a powerful alternative tool for the design and optimization of liquid-metal-cooled miniature heat sinks.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

22 extracted references · 18 canonical work pages

  1. [12]

    Enhancement of liquid sodium (na) forced convection within miniature heat sinks,

    M. Pourghasemi and N. Fathi, “Enhancement of liquid sodium (na) forced convection within miniature heat sinks,” ASME Journal of Heat and Mass Transfer, vol. 145, no. 5, 2023

    work page 2023

  2. [1]

    Sensitivity analysis and application of machine learning methods to predict the heat transfer performance of cnt/water nanofluid flows through coils,

    A. Baghban, M. Kahani, M. A. Nazari, M. H. Ahmadi, and W. -M. Yan, “Sensitivity analysis and application of machine learning methods to predict the heat transfer performance of cnt/water nanofluid flows through coils,” International Journal of Heat and Mass Transfer, vol. 128, pp. 825– 835, 2019

    work page 2019

  3. [2]

    Estimation of air dew point temperature using computational intelligence schemes,

    A. Baghban, M. Bahadori, J. Rozyn, M. Lee, A. Abbas, A. Bahadori, and A. Rahimali, “Estimation of air dew point temperature using computational intelligence schemes,” Applied thermal engineering, vol. 93, pp. 1043–1052, 2016

    work page 2016

  4. [3]

    Prediction of thermal conductivity of ethylene glycol –water solutions by using artificial neural networks,

    H. Kurt and M. Kayfeci, “Prediction of thermal conductivity of ethylene glycol –water solutions by using artificial neural networks,” Applied energy, vol. 86, no. 10, pp. 2244–2248, 2009

    work page 2009

  5. [4]

    Modeling viscosity of nanofluids using diffusional neural networks,

    F. Yousefi, H. Karimi, and M. M. Papari, “Modeling viscosity of nanofluids using diffusional neural networks,” Journal of Molecular Liquids, vol. 175, pp. 85–90, 2012

    work page 2012

  6. [5]

    Convolutional-network models to predict wall -bounded turbulence from wall quantities,

    L. Guastoni, A. Güemes, A. Ianiro, S. Discetti, P. Schlatter, H. Azizpour, and R. Vinuesa, “Convolutional-network models to predict wall -bounded turbulence from wall quantities,” Journal of Fluid Mechanics, vol. 928, p. A27, 2021

    work page 2021

  7. [6]

    A Review of Physics-Informed Machine Learning in Fluid Mechanics

    PINN-Sharma, Pushan, Wai Tong Chung, Bassem Akoush, and Matthias Ihme. "A Review of Physics-Informed Machine Learning in Fluid Mechanics." Energies 16, no. 5 (2023): 2343

    work page 2023

  8. [7]

    Finite volume method network for the acceleration of unsteady computational fluid dynamics: Non -reacting and reacting flows,

    J. Jeon, J. Lee, and S. J. Kim, “Finite volume method network for the acceleration of unsteady computational fluid dynamics: Non -reacting and reacting flows,” International Journal of Energy Research, vol. 46, no. 8, pp. 10770–10795, 2022

    work page 2022

Show all 22 references
  1. [8]

    Physics -informed transfer learning strategy to accelerate unsteady fluid flow simulations,

    J. Jeon, J. Lee, H. Eivazi, R. Vinuesa, and S. J. Kim, “Physics -informed transfer learning strategy to accelerate unsteady fluid flow simulations,” arXiv preprint arXiv:2206.06817, 2022

    2022 arXiv

  2. [9]

    A comprehensive survey on transfer learning,

    F. Zhuang, Z. Qi, K. Duan, D. Xi, Y. Zhu, H. Zhu, H. Xiong, and Q. He, “A comprehensive survey on transfer learning,” Proceedings of the IEEE, vol. 109, no. 1, pp. 43–76, 2020

    2020

  3. [10]

    Convolutional neural network- based transfer learning for optical aerial images change detection,

    J. Liu, K. Chen, G. Xu, X. Sun, M. Yan, W. Diao, and H. Han, “Convolutional neural network- based transfer learning for optical aerial images change detection,” IEEE Geoscience and Remote Sensing Letters, vol. 17, no. 1, pp. 127–131, 2019

    2019

  4. [11]

    Hierarchical intelligent real-time optimal control for llrf using time series machine learning methods and transfer learning,

    R. Pirayesh, S. Biedron et al., “Hierarchical intelligent real-time optimal control for llrf using time series machine learning methods and transfer learning,” 2021

    2021

  5. [13]

    Predictions of reynolds and nusselt numbers in turbulent convection using machine learning models,

    S. Bhattacharya, M. K. Verma, and A. Bhattacharya, “Predictions of reynolds and nusselt numbers in turbulent convection using machine learning models,” Physics of Fluids, vol. 34, no. 2, 2022

    2022

  6. [14]

    Artificial neural network modeling of nanofluid flow in a microchannel heat sink using experimental data,

    M. M. Tafarroj, O. Mahian, A. Kasaeian, K. Sakamatapan, A. S. Dalkilic, and S. Wongwises, “Artificial neural network modeling of nanofluid flow in a microchannel heat sink using experimental data,” International Communications in Heat and Mass Transfer, vol. 86, pp. 25– 31, 2017

    2017

  7. [15]

    Monte Carlo cross -validation for a study with binary outcome and limited sample size

    Shan, Guogen. "Monte Carlo cross -validation for a study with binary outcome and limited sample size." BMC Medical Informatics and Decision Making 22, no. 1 (2022): 1-15

    2022

  8. [16]

    Triaging patient complaints: Monte Carlo cross-validation of six machine learning classifiers

    Elmessiry, Adel, William O. Cooper, Thomas F. Catron, Jan Karrass, Zhe Zhang, and Munindar P. Singh. "Triaging patient complaints: Monte Carlo cross-validation of six machine learning classifiers." JMIR medical informatics 5, no. 3 (2017): e7140

    2017

  9. [17]

    Computational framework development for heat transfer studies in liquid metal-cooled small-scale heat sinks with non-circular cross-sections,

    M. Pourghasemi, and N. Fathi, “Computational framework development for heat transfer studies in liquid metal-cooled small-scale heat sinks with non-circular cross-sections,” Journal of Thermal Science and Engineering Applications, pp.1-16, 2023

    2023

  10. [18]

    Designing monte carlo simulation and an optimal machine learning to optimize and model space missions,

    R. Pirayeshshirazinezhad, S. G. Biedron, J. A. D. Cruz, S. S. Güitrón, and´ M. Martínez - Ramón, “Designing monte carlo simulation and an optimal machine learning to optimize and model space missions,” IEEE Access, vol. 10, pp. 45643–45662, 2022

    2022

  11. [19]

    Artificial intelligence, controls, and sensor fusion for optimization and modeling of space missions and particle accelerators,

    R. Pirayeshshirazinezhad, “Artificial intelligence, controls, and sensor fusion for optimization and modeling of space missions and particle accelerators,” 2022

    2022

  12. [20]

    Shawe -Taylor, N

    J. Shawe -Taylor, N. Cristianini et al., Kernel methods for pattern analysis. Cambridge university press, 2004

    2004

  13. [21]

    Variations of box plots,

    R. McGill, J. W. Tukey, and W. A. Larsen, “Variations of box plots,” The american statistician, vol. 32, no. 1, pp. 12–16, 1978

    1978

  14. [22]

    On a test of whether one of two random variables is stochastically larger than the other,

    H. B. Mann and D. R. Whitney, “On a test of whether one of two random variables is stochastically larger than the other,” The annals of mathematical statistics, pp. 50–60, 1947

    1947

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.