Pith. sign in

REVIEW 3 major objections 4 minor 40 references

A single machine-learning model predicts the sputtering parameters needed to deposit a target alloy composition, and recovers composition from process settings.

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

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

T0 review · deepseek-v4-flash

2026-08-04 18:31 UTC pith:HX7LIKVA

load-bearing objection A plausible in-sample proof of concept for ML-guided composition control in multicathode PVD; the element-independence claim outruns the evidence. the 3 major comments →

arxiv 2608.01903 v1 pith:HX7LIKVA submitted 2026-08-03 cond-mat.mtrl-sci

Machine Learning Assisted Design of Complex and High Entropy Alloys by Hybrid HiPIMS/Pulsed-DC PVD Process for Low Carbon Energy Applications in Extreme Environments

Paul Foulquier , Ryma Haddad , Ali Mahmoud , Eric Monsifrot , Fanny Balbaud-Celerier , Jean-Philippe Poli , Frederic Schuster This is my paper
classification cond-mat.mtrl-sci
keywords high-entropy alloyscomplex concentrated alloyscombinatorial magnetron sputteringinverse designmachine learningelement-independent modelsputtering yieldnuclear coatings
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper claims that a generic, element-independent machine learning model can replace the usual trial-and-error loop in combinatorial magnetron sputtering: given the desired chemical composition of a complex or high-entropy alloy coating, it predicts the power to apply on each cathode, and given the powers it predicts the resulting composition. The model uses physical descriptors such as sputtering yield and cathode operating mode instead of element identity, so it should transfer to new alloy systems without retraining on a new dataset. The authors demonstrate this on a 7-element, 82-point dataset with forward prediction error around 0.1 atomic fraction and backward prediction error around 70 W. If correct, this would sharply cut the experimental effort in exploring complex compositional spaces for nuclear and low-carbon energy coatings.

Core claim

The central claim is the feasibility of a first prototype generic machine learning framework for inverse design in hybrid HiPIMS/pulsed-DC PVD. Rather than training one model per alloy system, the authors encode each cathode by its operating mode and each element by normalized physical parameters, notably sputtering yield. On this basis, Random Forest and XGBoost models learn both the forward map (power -> composition) and the backward map (composition -> power), the latter regularized by bounding the predicted power windows to handle its non-injective nature. The model is validated on a dataset of 82 depositions spanning seven elements (Cr, Al, Mo, Ni, V, W, Nb), with correlations between p

What carries the argument

The element-independent descriptor encoding: each deposition event is described by the cathode operating mode (HiPIMS or pulsed-DC) and normalized physical parameters of the sputtered element, chiefly the sputtering yield, instead of chemical identity. This is what lets a single Random Forest/XGBoost model serve any alloy system; the inverse problem is made tractable by predicting power within user-specified windows, which bounds the otherwise infinite set of power combinations giving the same composition.

Load-bearing premise

The model's transferability rests on the assumption that cathode operating mode plus normalized sputtering yield capture enough of the deposition physics to predict any element combination, yet no element is held out from the training set to verify this.

What would settle it

Train the element-independent model on any six of the seven elements, then use it to predict the cathode powers for a target composition containing only the held-out seventh element, deposit that coating, and measure the actual composition; if the error is comparable to or larger than the reported ~0.1 RMSE rather than the in-distribution error, the element-independence claim fails.

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

Share X Bluesky LinkedIn Reddit HN

If this is right

  • New alloy systems can be explored by swapping target elements and reusing the same model, requiring only a handful of calibration depositions rather than a full retraining campaign.
  • The inverse map gives experimentalists a concrete starting recipe for a target composition, reducing the number of deposition runs needed to land on a specified high-entropy alloy.
  • The correlation matrices show the model captures cross-cathode plasma interactions, so it can be used to anticipate how changing one cathode's power perturbs the whole film composition.
  • The same descriptor logic can be extended to predict microstructural and functional properties once grain size, texture, stress, and deposition rate are added as targets.
  • The approach supports the long-term goal of closed-loop autonomous deposition, where in-situ diagnostics feed corrections back into the model in real time.

Where Pith is reading between the lines

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

  • The strongest test of element-independence would be a leave-one-element-out experiment: train on six elements and ask the model to predict the power for a seventh held-out element; the paper does not report this, so transferability to truly unseen elements remains an open claim.
  • Sputtering yield alone may not separate elements whose yields are similar but whose plasma chemistry differs (e.g., target poisoning, secondary electron emission, gas rarefaction); adding plasma diagnostics as descriptors could make the model robust to such cases.
  • The same descriptor-based inverse-design approach could be carried over to reactive sputtering of oxides or nitrides, where composition targets become stoichiometry targets and the descriptor set would need to include reactive gas flow and target state.
  • Because the backward map is surjective, the model as presented gives one recipe among many; a human or optimisation loop must impose additional constraints (cost, rate, stress) to select the best recipe.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper presents a combination of a programmatic overview of the French DIADEM materials-data-science initiative and a machine-learning study on the DIADEM-2D combinatorial magnetron sputtering platform. The authors describe the selection of candidate alloy systems for molten-salt corrosion (Ni-Cr-Mo-Al) and nuclear cladding diffusion barriers (V-Nb-Mo-W), then train Random Forest, XGBoost, Bayesian Ridge, and a fuzzy inference system on a dataset of seven elements (Ni, Cr, Mo, Al, W, Nb, V; 82 points). The forward task predicts coating composition from cathode powers; the inverse task predicts powers from a target composition. The authors report RMSE ≈ 0.1 (forward) and RMSE ≈ 70 W (backward) and claim that, because the model uses physical descriptors such as sputtering yield rather than element identity, it is element-independent and transferable to new elements without retraining. The central conclusion is that this is a first prototype of a generic machine-learning framework for inverse design of PVD process parameters.

Significance. If the element-transferability claim were established, this would be a valuable contribution to high-throughput combinatorial sputtering, substantially reducing experimental effort when exploring new alloy systems. The paper has clear strengths: it describes a concrete closed-loop platform concept, checks model behavior through correlation matrices, and explores interpretable fuzzy systems in addition to black-box ensembles. However, the evidence presented is an in-sample proof of concept only. The evaluation uses random splits in which every element appears in training, no element is held out, there is no comparison against a simple physical scaling baseline, and no uncertainty quantification or per-element errors are reported. The central 'generic, element-independent' claim is therefore not yet supported by the data. The work could become a meaningful demonstration if the authors add leave-one-element-out cross-validation, per-element and test-only metrics, a descriptor-based baseline, and public data/code.

major comments (3)
  1. [§4.2, §4.3, §4.4] The load-bearing claim is element independence and transferability, but it is never tested. Section 4.3 describes a random 80/20 split repeated five times; in such a split every element appears in the training set, so the test set only measures interpolation among already seen elements. Section 4.4 then states that 'This methodology can be applied to dataset with elements which are not present in the dataset', a claim not supported by any experiment. With only seven elements, descriptors such as sputtering yield are nearly unique per element and may serve as an implicit element label. Please add leave-one-element-out cross-validation, report per-element errors, and compare against a simple sputtering-yield normalization baseline (e.g., composition approximately proportional to yield × power normalized over cathodes). Without such a test, the Conclusion's transferability statement is unsu
  2. [§4.3, §4.4, Figures 10, 11] The reported accuracy lacks the quantitative detail needed to assess the claims. The parity plots appear to show train and test points together, and the RMSE values (≈0.1 and ≈70 W) are given without normalization or scale context: what is the composition unit (mole fraction?) and what is the range of powers in the dataset? The repeated 5-fold split provides only an average; no error bars, confidence intervals, or per-element residuals are shown. Please report separate train/test RMSE with standard deviations, describe the composition and power ranges, and plot test points separately from training points. This is necessary for the 'accuracy' claim in the abstract and conclusion.
  3. [§4.2] The element-independent descriptor set is under-specified. The text names 'cathode operating mode' and 'sputtering yield' but does not give the full feature vector, the energy at which sputtering yield is evaluated, or an explicit statement that categorical element identity variables are excluded. Without this information the model is not reproducible, and the claim of element independence cannot be checked. Please provide a complete descriptor table, the normalization procedure, and a code/data release. A helpful additional check would be to train on a subset of elements and show that the learned mapping remains physically sensible for a held-out element, even if the eventual target is one of the seven studied elements.
minor comments (4)
  1. [§II] Typo: 'Thises ways' should be 'These ways'. Also 'reliance' in §4.3 should be 'reliability'.
  2. [§4.4, Figures 10, 11] The phrase 'test/train graph' or 'test/train batch' is misleading. Please distinguish training and test points clearly, for example with different markers and a separate test-only panel, and state the number of points in each set.
  3. [Figure 14] Axes of the learning-curve figure are unlabeled. Define CV RMSE in the caption and indicate which line corresponds to which model.
  4. [Data availability] 'Data are available from the corresponding author upon reasonable request' is not sufficient for reproducibility of the ML results. Please deposit the dataset and analysis code in a public repository, or specify any legal/ethical restrictions preventing release.

Circularity Check

0 steps flagged

No significant circularity: the ML predictions are empirical fits with random-split test-set evaluation; the transferability claim is an untested extrapolation, not a construction-equivalent reduction.

full rationale

The paper's central results are regression models trained on 82 experimental points and evaluated on random 80/20 splits with 5-fold repetition (Section 4.3). A random-split test set is a standard internal generalization check, so the reported RMSE values are not fitted parameters renamed as predictions. The 'element-independent' claim in Section 4.2 rests on the use of physical descriptors such as sputtering yield instead of element names; this is a representational choice, not an equation that makes the output equal to the input. The statement in Section 4.4 that the methodology 'can be applied to dataset with elements which are not present in the dataset' is an extrapolation that is never tested with a leave-one-element-out experiment; that is a validity/robustness gap, not circular reasoning. Self-citations [4], [23], [38], [39] are background or auxiliary and do not carry the central predictive claim. Therefore the derivation chain is not circular, though the transferability claim is weaker than demonstrated.

Axiom & Free-Parameter Ledger

2 free parameters · 3 axioms · 0 invented entities

The central empirical claim is an interpolative ML fit; it introduces no new physics. Its transferability rests on an unverified sufficiency of physical descriptors, and the fitted model parameters are not reported. No independent external dataset or held-out element is used.

free parameters (2)
  • Machine-learning model parameters and hyperparameters (Random Forest, XGBoost, Bayesian ridge) = not reported in article
    RF/XGBoost ensemble parameters and Optuna-tuned hyperparameters are fitted to the 82-point dataset; no final values or seeds are given (Section 4.3).
  • Inverse-model power windows = not specified
    Section 4.4 introduces manually bounded windows to make the backward composition-to-power mapping well-posed; the window bounds are a hand-chosen modeling input.
axioms (3)
  • domain assumption Cathode power and tabulated sputtering yield, plus cathode operating mode, are sufficient descriptors to predict composition across elements.
    Section 4.2 states the model uses these descriptors rather than chemical identity; the claim of element independence rests on this, but no held-out element validation is provided.
  • domain assumption All deposition conditions other than cathode powers are fixed and can be neglected in the mapping.
    Section II fixes Ar pressure, bias, etching, and pulse parameters; the model then treats composition as a function of power only, ignoring possible run-to-run drift and plasma interactions not captured by descriptors.
  • domain assumption Random 80/20 splits of the 82 datapoints are representative and the five-fold CV estimates generalization.
    Section 4.3 uses random splits and repeated cross-validation on a small dataset; this assumes no strong temporal or compositional leakage.

pith-pipeline@v1.3.0-daily-deepseek · 10749 in / 10032 out tokens · 114253 ms · 2026-08-04T18:31:28.474644+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of Machine Learning Assisted Design of Complex and High Entropy Alloys by Hybrid HiPIMS/Pulsed-DC PVD Process for Low Carbon Energy Applications in Extreme Environments." pith.science (2026). https://pith.science/paper/HX7LIKVA

@misc{pith2026260801903,
  author       = {Pith},
  title        = {Pith review of: Machine Learning Assisted Design of Complex and High Entropy Alloys by Hybrid HiPIMS/Pulsed-DC PVD Process for Low Carbon Energy Applications in Extreme Environments},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HX7LIKVA}},
  note         = {Machine review of arXiv:2608.01903}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Complex and high entropy alloys are attracting much attention currently thanks to their mechanical and corrosion resistance properties in harsh environments, in particular needed for carbon-free energy applications. However, their elaboration in bulk and in thin film form in a trial-and-error approach is impractical due to their complexity and the cocktail effect. The recent development of artificial intelligence brings a new possibility for their elaboration and adjustment of their properties. Firstly, we present an overview of Materials and data science research. Then we describe how DIADEM - French initiative for Materials and Data science convergence - tackles the development of innovative coatings for carbon-free energy applications (nuclear, high temperature electrolysis, ...) thanks to the development of a nationwide network of synthesis and characterization platforms - the DIADEM discovery hub. We describe in particular DIADEM-2D, an AI-driven Hybrid HiPIMS/Pulsed-DC PVD process using 4 cathodes in confocal combinatorial configuration. We present the high entropy alloy determination using data from the literature for corrosion resistance in molten salt media and nuclear accidental conditions. An element-independent model gathering deposition parameters and coating properties has been implemented allowing the design of protective coatings with a particular composition. We demonstrate the feasibility of this process and its accuracy.

Figures

Figures reproduced from arXiv: 2608.01903 by Ali Mahmoud, Eric Monsifrot, Fanny Balbaud-Celerier, Frederic Schuster, Jean-Philippe Poli, Paul Foulquier, Ryma Haddad.

Figure 1
Figure 1. Figure 1: Worldwide R&D dynamics in Data and Materials science convergence over a 30 years period starting in 1996 with [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Academics dynamics per geographical region over a 10 years period starting in 2015 showing an increase of China [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: European R&D dynamics in data and materials science convergence over a 10 years period starting in 2016. France [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: a) DIADEM-2D platform at INSTN-Saclay, b) Confocal combinatorial deposition as fitted in DIADEM-2D, c) [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Operation description of DIADEM-2D a) to date, b) in a near future including an AI-controlled feedback loop for [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: a) Alloy ranking analysed in molten salt corrosion experiments, b) Comparison of scores computed by Spring Rank [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: a) Cr-Zr phase diagram, b) SEM image with EDS profile of a Cr-coated Zr alloy cladding displaying bloating and Cr [PITH_FULL_IMAGE:figures/full_fig_p007_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: a) Architecture of a Cr + diffusion barrier coated Zr alloy nuclear cladding as planned in the ASTERIX project, b) [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Sketch of the learning process used in this study using test and train batches, optuna optimizer and cross [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Test/train batch using the element-independent model with Random Forest and XGBoost in the forward direction [PITH_FULL_IMAGE:figures/full_fig_p010_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Test/train batch using the element-independent model with Random Forest and XGBoost in the backward [PITH_FULL_IMAGE:figures/full_fig_p011_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Pearson and Spearman experimental correlation matrices calculated using data from the experimental database [PITH_FULL_IMAGE:figures/full_fig_p012_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Forward and backward correlation matrices using experimental data and predictions using Random forest, [PITH_FULL_IMAGE:figures/full_fig_p012_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Learning curves displaying training RMSE and CV RMSE for the calculations of Random Forest, XGBoost and [PITH_FULL_IMAGE:figures/full_fig_p013_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Test/train batch using the element-independent model with a Fuzzy Inference System. On the left-hand side, the [PITH_FULL_IMAGE:figures/full_fig_p013_15.png] view at source ↗

discussion (0)

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

Reference graph

Works this paper leans on

40 extracted references · 19 canonical work pages

  1. [1]

    De Goes Foschiani, S

    B. De Goes Foschiani, S. Bongiorno, O. Robach, O. Ulrich, J. Micha, et J. Eymery, « Multimodal Imaging of Strain and Light Emission of Core–Shell InGaN/GaN Wires under a Sub-Micrometer Polychromatic X-Ray Probe », Adv. Mater. Interfaces, vol. 13, no 4, p. e01025, févr. 2026, doi: 10.1002/admi.202501025

    work page doi:10.1002/admi.202501025 2026

  2. [2]

    J. J. De Pablo, B. Jones, C. L. Kovacs, V. Ozolins, et A. P. Ramirez, « The Materials Genome Initiative, the interplay of experiment, theory and computation », Curr. Opin. Solid State Mater. Sci., vol. 18, no 2, p. 99‑117, avr. 2014, doi: 10.1016/j.cossms.2014.02.003

    work page doi:10.1016/j.cossms.2014.02.003 2014

  3. [3]

    Bayerlein et al., « Concepts for a Semantically Accessible Materials Data Space: Overview over Specific Implementations in Materials Science », Adv

    B. Bayerlein et al., « Concepts for a Semantically Accessible Materials Data Space: Overview over Specific Implementations in Materials Science », Adv. Eng. Mater., vol. 27, no 8, p. 2401092, avr. 2025, doi: 10.1002/adem.202401092

    work page doi:10.1002/adem.202401092 2025

  4. [4]

    Lomello, L

    F. Lomello, L. Bard, M. Maglione, et F. Schuster, « PEPR DIADEM: Priority equipment and research program on the development of innovative materials using artificial intelligence », Comput. Struct. Biotechnol. J., vol. 25, p. 186‑193, déc. 2024, doi: 10.1016/j.csbj.2024.09.019

    work page doi:10.1016/j.csbj.2024.09.019 2024

  5. [5]

    Cantor, I

    B. Cantor, I. T. H. Chang, P. Knight, et A. J. B. Vincent, « Microstructural development in equiatomic multicomponent alloys », Mater. Sci. Eng. A, vol. 375‑377, p. 213‑218, juill. 2004, doi: 10.1016/j.msea.2003.10.257

    doi:10.1016/j.msea.2003.10.257 2004

  6. [6]

    Yeh, « Alloy Design Strategies and Future Trends in High-Entropy Alloys », JOM, vol

    J.-W. Yeh, « Alloy Design Strategies and Future Trends in High-Entropy Alloys », JOM, vol. 65, no 12, p. 1759‑1771, déc. 2013, doi: 10.1007/s11837-013-0761-6

    work page doi:10.1007/s11837-013-0761-6 2013

  7. [7]

    Y. Y. Chen, T. Duval, U. D. Hung, J. W. Yeh, et H. C. Shih, « Microstructure and electrochemical properties of high entropy alloys—a comparison with type-304 stainless steel », Corros. Sci., vol. 47, no 9, p. 2257‑2279, sept. 2005, doi: 10.1016/j.corsci.2004.11.008

    work page doi:10.1016/j.corsci.2004.11.008 2005

  8. [8]

    Y. Y. Chen, U. T. Hong, H. C. Shih, J. W. Yeh, et T. Duval, « Electrochemical kinetics of the high entropy alloys in aqueous environments—a comparison with type 304 stainless steel », Corros. Sci., vol. 47, no 11, p. 2679‑2699, nov. 2005, doi: 10.1016/j.corsci.2004.09.026

    work page doi:10.1016/j.corsci.2004.09.026 2005

  9. [9]

    J. -W. Yeh et al., « Nanostructured High-Entropy Alloys with Multiple Principal Elements: Novel Alloy Design Concepts and Outcomes », Adv. Eng. Mater., vol. 6, no 5, p. 299‑303, mai 2004, doi: 10.1002/adem.200300567

    doi:10.1002/adem.200300567 2004

  10. [10]

    C. M. Rost et al., « Entropy-stabilized oxides », Nat. Commun., vol. 6, no 1, p. 8485, sept. 2015, doi: 10.1038/ncomms9485

    doi:10.1038/ncomms9485 2015

  11. [11]

    S. S. I. Almishal et al., « Untangling individual cation roles in rock salt high-entropy oxides », Acta Mater., vol. 279, p. 120289, oct. 2024, doi: 10.1016/j.actamat.2024.120289

    arXiv 2024

  12. [12]

    Su et al., « Visualizing the Formation of High-Entropy Fluorite Oxides from an Amorphous Precursor at Atomic Resolution », ACS Nano, vol

    L. Su et al., « Visualizing the Formation of High-Entropy Fluorite Oxides from an Amorphous Precursor at Atomic Resolution », ACS Nano, vol. 16, no 12, p. 21397‑21406, déc. 2022, doi: 10.1021/acsnano.2c09760

    work page doi:10.1021/acsnano.2c09760 2022

  13. [13]

    G. H. J. Johnstone et al., « Entropy Engineering and Tunable Magnetic Order in the Spinel High- Entropy Oxide », J. Am. Chem. Soc., vol. 144, no 45, p. 20590‑20600, nov. 2022, doi: 10.1021/jacs.2c06768

    work page doi:10.1021/jacs.2c06768 2022

  14. [14]

    Teng et al., « Synthesis and structures of high-entropy pyrochlore oxides », J

    Z. Teng et al., « Synthesis and structures of high-entropy pyrochlore oxides », J. Eur. Ceram. Soc., vol. 40, no 4, p. 1639‑1643, avr. 2020, doi: 10.1016/j.jeurceramsoc.2019.12.008

    work page doi:10.1016/j.jeurceramsoc.2019.12.008 2020

  15. [15]

    D. A. Vinnik et al., « High-entropy oxide phases with magnetoplumbite structure », Ceram. Int., vol. 45, no 10, p. 12942‑12948, juill. 2019, doi: 10.1016/j.ceramint.2019.03.221

    work page doi:10.1016/j.ceramint.2019.03.221 2019

  16. [16]

    J. Ma, T. Liu, W. Ye, Q. He, et K. Chen, « High-entropy perovskite oxides for energy materials: A review », J. Energy Storage, vol. 90, p. 111890, juin 2024, doi: 10.1016/j.est.2024.111890

    arXiv 2024

  17. [17]

    Feng et al., « Discovery of high entropy garnet solid-state electrolytes via ultrafast synthesis », Energy Storage Mater., vol

    Y. Feng et al., « Discovery of high entropy garnet solid-state electrolytes via ultrafast synthesis », Energy Storage Mater., vol. 63, p. 103053, nov. 2023, doi: 10.1016/j.ensm.2023.103053

    arXiv 2023

  18. [18]

    forward” direction (power  composition) has a bijective character, the “backward

    with a wide variety of properties which make them interesting for different applications like thermoelectricity, magnetism, catalysis or batteries [19]. High entropy materials are subject to cocktail effect [20] stating that the material structure and properties are determined by the combination of the properties of the elements and the interactions betwe...

  19. [19]

    X. Miao, Z. Peng, L. Shi, et S. Zhou, « Insulating High-Entropy Ruthenium Oxide as a Highly Efficient Oxygen-Evolving Electrocatalyst in Acid », ACS Catal., vol. 13, no 6, p. 3983‑3989, mars 2023, doi: 10.1021/acscatal.2c06276. 17

    work page doi:10.1021/acscatal.2c06276 2023

  20. [20]

    S. Sen, M. Palabathuni, K. M. Ryan, et S. Singh, « High Entropy Oxides: Mapping the Landscape from Fundamentals to Future Vistas: Focus Review », ACS Energy Lett., vol. 9, no 8, p. 3694‑3718, août 2024, doi: 10.1021/acsenergylett.4c01129

    doi:10.1021/acsenergylett.4c01129 2024

  21. [21]

    Yeh, « Recent progress in high-entropy alloys », Ann

    J.-W. Yeh, « Recent progress in high-entropy alloys », Ann. Chim. Sci. Matér., vol. 31, no 6, p. 633‑648, déc. 2006, doi: 10.3166/acsm.31.633-648

    work page doi:10.3166/acsm.31.633-648 2006

  22. [22]

    Ranganathan, « Alloyed pleasures: Multimetallic cocktails », Curr

    S. Ranganathan, « Alloyed pleasures: Multimetallic cocktails », Curr. Sci., vol. 85, no 10, 2003

    2003

  23. [23]

    Kouznetsov, K

    V. Kouznetsov, K. Macák, J. M. Schneider, U. Helmersson, et I. Petrov, « A novel pulsed magnetron sputter technique utilizing very high target power densities », Surf. Coat. Technol., vol. 122, no 2‑3, p. 290‑293, déc. 1999, doi: 10.1016/S0257-8972(99)00292-3

    work page doi:10.1016/s0257-8972(99)00292-3 1999

  24. [24]

    Herschberg et al., « From Pairwise Comparisons of Complex Behavior to an Overall Performance Rank: A Novel Alloy Design Strategy », Metals, vol

    R. Herschberg et al., « From Pairwise Comparisons of Complex Behavior to an Overall Performance Rank: A Novel Alloy Design Strategy », Metals, vol. 14, no 12, p. 1412, déc. 2024, doi: 10.3390/met14121412

    work page doi:10.3390/met14121412 2024

  25. [25]

    Brachet et al., « Gaines de combustible nucléaire, procédé de fabrication et utilisations contre l’oxydation/hydruration », EP3195322B1, 26 juillet 2017

    J.-C. Brachet et al., « Gaines de combustible nucléaire, procédé de fabrication et utilisations contre l’oxydation/hydruration », EP3195322B1, 26 juillet 2017

    2017

  26. [26]

    Krejčí et al., « Development and testing of multicomponent fuel cladding with enhanced accidental performance », Nucl

    J. Krejčí et al., « Development and testing of multicomponent fuel cladding with enhanced accidental performance », Nucl. Eng. Technol., vol. 52, no 3, p. 597‑609, mars 2020, doi: 10.1016/j.net.2019.08.015

    work page doi:10.1016/j.net.2019.08.015 2020

  27. [27]

    Okamoto, « Supplemental Literature Review of Binary Phase Diagrams: B-Fe, Cr-Zr, Fe-Np, Fe- W, Fe-Zn, Ge-Ni, La-Sn, La-Ti, La-Zr, Li-Sn, Mn-S, and Nb-Re », J

    H. Okamoto, « Supplemental Literature Review of Binary Phase Diagrams: B-Fe, Cr-Zr, Fe-Np, Fe- W, Fe-Zn, Ge-Ni, La-Sn, La-Ti, La-Zr, Li-Sn, Mn-S, and Nb-Re », J. Phase Equilibria Diffus., vol. 37, no 5, p. 621‑634, oct. 2016, doi: 10.1007/s11669-016-0465-z

    work page doi:10.1007/s11669-016-0465-z 2016

  28. [28]

    B. Wang, Q. Wang, N. Lu, X. Liang, et B. Shen, « Enhanced high-temperature strength of HfNbTaTiZrV refractory high-entropy alloy via Al2O3 reinforcement », J. Mater. Sci. Technol., vol. 123, p. 191‑200, oct. 2022, doi: 10.1016/j.jmst.2022.01.025

    work page doi:10.1016/j.jmst.2022.01.025 2022

  29. [29]

    D. B. Miracle, M.-H. Tsai, O. N. Senkov, V. Soni, et R. Banerjee, « Refractory high entropy superalloys (RSAs) », Scr. Mater., vol. 187, p. 445‑452, oct. 2020, doi: 10.1016/j.scriptamat.2020.06.048

    doi:10.1016/j.scriptamat.2020.06.048 2020

  30. [30]

    Isayev et P

    R. Isayev et P. Dzhumaev, « Interaction of a diffusion barrier from the refractory metals with a zirconium alloy and a chrome coating of an accident tolerant fuel », Nucl. Eng. Des., vol. 407, p. 112307, juin 2023, doi: 10.1016/j.nucengdes.2023.112307

    arXiv 2023

  31. [31]

    Lee et al., « An experimentally driven high-throughput approach to design refractory high- entropy alloys », Mater

    C. Lee et al., « An experimentally driven high-throughput approach to design refractory high- entropy alloys », Mater. Des., vol. 223, p. 111259, nov. 2022, doi: 10.1016/j.matdes.2022.111259

    arXiv 2022

  32. [32]

    Pedregosa et al., « Scikit-learn: Machine Learning in Python », 2012, doi: 10.48550/ARXIV.1201.0490

    F. Pedregosa et al., « Scikit-learn: Machine Learning in Python », 2012, doi: 10.48550/ARXIV.1201.0490

    Pith/arXiv arXiv doi:10.48550/arxiv.1201.0490 2012

  33. [33]

    Breiman, « Random Forests », Mach

    L. Breiman, « Random Forests », Mach. Learn., vol. 45, no 1, p. 5‑32, oct. 2001, doi: 10.1023/A:1010933404324

    doi:10.1023/a:1010933404324 2001

  34. [34]

    Chen et C

    T. Chen et C. Guestrin, « XGBoost: A Scalable Tree Boosting System », in Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco California USA: ACM, août 2016, p. 785‑794. doi: 10.1145/2939672.2939785

    arXiv 2016

  35. [35]

    Akiba, S

    T. Akiba, S. Sano, T. Yanase, T. Ohta, et M. Koyama, « Optuna: A Next-generation Hyperparameter Optimization Framework », 2019, arXiv. doi: 10.48550/ARXIV.1907.10902

    Pith/arXiv arXiv doi:10.48550/arxiv.1907.10902 2019

  36. [36]

    Bates, T

    S. Bates, T. Hastie, et R. Tibshirani, « Cross-Validation: What Does It Estimate and How Well Does It Do It? », J. Am. Stat. Assoc., vol. 119, no 546, p. 1434‑1445, avr. 2024, doi: 10.1080/01621459.2023.2197686

    arXiv 2024

  37. [37]

    D. J. C. MacKay, « Bayesian Interpolation », Neural Comput., vol. 4, no 3, p. 415‑447, mai 1992, doi: 10.1162/neco.1992.4.3.415

    doi:10.1162/neco.1992.4.3.415 1992

  38. [38]

    Tipping, « Sparse Bayesian Learning and the Relevance Vector Machine », J

    M. Tipping, « Sparse Bayesian Learning and the Relevance Vector Machine », J. Mach. Learn. Res., no 1, p. 211‑244, 2001

    2001

  39. [39]

    Hajri, J.-P

    H. Hajri, J.-P. Poli, et L. Boudet, « Towards Monotonous Functions Approximation from Few Data With Gradual Generalized Modus Ponens: Application to Materials Science », in 2021 IEEE 33rd International Conference on Tools with Artificial Intelligence (ICTAI), Washington, DC, USA: IEEE, nov. 2021, p. 796‑800. doi: 10.1109/ICTAI52525.2021.00127. 18

    arXiv 2021

  40. [40]

    Rousselle, J.-P

    O. Rousselle, J.-P. Poli, et N. B. Abdallah, « Towards an Interpretable Fuzzy Approach to Experimental Design », in Information Processing and Management of Uncertainty in Knowledge- Based Systems, vol. 1174, M.-J. Lesot, S. Vieira, M. Z. Reformat, J. P. Carvalho, F. Batista, B. Bouchon-Meunier, et R. R. Yager, Éd., in Lecture Notes in Networks and System...

    work page doi:10.1007/978-3-031-74003-9_18 2024