Pith. sign in

REVIEW 3 major objections 6 minor 35 references

Autonomous materials search using machine learning and ab initio calculations for L10-FePt-based quaternary alloys

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read An autonomous 100-day search over roughly 200,000 possible L10-FePt quaternary compositions converges on Fe1-xMnxPt1-yEry as a candidate whose computed magnetic moment and magnetocrystalline anisotropy both exceed those of FePt.

desk verdict A solid computational screening study that proposes Fe1-xMnxPt1-yEry as a T=0 KKR-CPA prediction with high M and EMCA, honestly caveated by the authors themselves; the practical room-temperature claim is not established. read the letter →

arxiv 2411.18907 v1 pith:2E344BFL submitted 2024-11-28 cond-mat.mtrl-sci

Yuma Iwasaki , Daisuke Ogawa , Masato Kotsugi , Yukiko K. Takahashi This is my paper · ORCID
classification cond-mat.mtrl-sci PACS 75.30.Gw75.50.Bb71.15.Mb
keywords L10-FePtquaternaryalloysautonomousmaterialssearchmachinelearningBayesianoptimizationmagnetocrystallineanisotropymagneticmomentKKR-CPA
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

To find quaternary L10-FePt-based alloys with better magnetic properties, the paper runs an autonomous loop that alternates ab initio calculations with machine-learning suggestions. The loop explored a composition space of roughly 200,000 candidates and, over 100 days, kept returning to Fe$_{1-x}$Mn$_x$Pt$_{1-y}$Er$_y$, which the authors propose as a candidate with both high magnetic moment $M$ and high magnetocrystalline anisotropy energy $E_{\rm MCA}$. The practical driver is magnetic recording: next-generation media need materials stronger than FePt on both properties, and exhaustive ab initio screening would be too expensive. The paper is explicit that its proposal is a speculative prediction pending experimental synthesis and measurement, and it flags a room-temperature caveat for the erbium moment. If the prediction survives, the demonstration matters because the same loop transfers to other large alloy spaces.

What carries the argument

The machinery is a closed loop. KKR-CPA (Korringa-Kohn-Rostoker coherent potential approximation, a Green's-function-based density-functional method for disordered alloys) computes $M$ and $E_{\rm MCA}$ for candidate compositions. A Gaussian-process Bayesian optimizer, using the upper confidence bound as its acquisition function, selects the candidate that most improves the Pareto hypervolume in the ($M$, $E_{\rm MCA}$) plane. An autoencoder compresses composition plus Magpie descriptors into a 30-dimensional latent vector that serves as the search space. On the physics side, the anisotropy gain is traced to the orbital moment anisotropy $\Delta M_o^T$, the difference in total orbital moment between the [001] and [100] magnetization directions; Er's large orbital moment, plus Er-induced increases in Fe, Mn, and Pt orbital moments, drives $E_{\rm MCA}$ upward.

What would settle it

Synthesize Fe$_{1-x}$Mn$_x$Pt$_{1-y}$Er$_y$ (for example $x, y$ near 0.1--0.2) as an ordered L10 film, measure its saturation magnetization and uniaxial magnetocrystalline anisotropy at room temperature, and compare with L10-FePt measured identically; if neither property exceeds FePt's, or if the erbium contribution to $M$ collapses as the paper's own caveat predicts, the central practical claim fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that autonomous exploration, rather than human intuition, identified Fe$_{1-x}$Mn$_x$Pt$_{1-y}$Er$_y$ as an L10 (ordered tetragonal) quaternary system in which both design targets---total magnetic moment $M$ and magnetocrystalline anisotropy energy $E_{\rm MCA}$---exceed the starting FePt values over a wide composition range. In the KKR-CPA calculations, adding Mn raises the Fe local moment and contributes its own large moment, while adding Er replaces low-moment Pt with a much larger moment, increasing $M$; meanwhile $E_{\rm MCA}$ increases because Er's large orbital moment and the induced increases in Fe, Mn, and Pt orbital moments enlarge the orbital moment anisotropy $\Delta M_o^T$. The paper states this as a successful demonstration of simulation-based autonomous search for quaternary L10 alloys, and immediately notes that the alloys are speculative predictions whose accuracy and structural stability have not been verified.

Load-bearing premise

The load-bearing premise is that the zero-temperature KKR-CPA results are representative at room temperature; the paper itself warns that erbium's moment, which drives much of the magnetization gain, decreases significantly at room temperature.

Editorial extensions

If this is right

  • If correct, Fe$_{1-x}$Mn$_x$Pt$_{1-y}$Er$_y$ becomes a concrete composition to try for next-generation magnetic recording, since it is computed to exceed FePt on both $M$ and $E_{\rm MCA}$.
  • The same closed loop can be applied to other quaternary or higher-order alloy families where exhaustive ab initio evaluation is financially or computationally prohibitive.
  • The Pareto-hypervolume selection rule shows how a search can optimize two competing properties at once, so the method transfers to any pair of computable materials properties.
  • The identification of $\Delta M_o^T$ as the anisotropy driver gives a screening principle: prioritize elements that enlarge the orbital moment along the easy axis.

Reading between the lines

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

  • A finite-temperature version of the loop is the natural next step; the paper does not test whether room-temperature magnetic moments would still put FeMnPtEr ahead of FePt.
  • The near-cubic $c/a$ at high Er content is left unresolved; because anisotropy typically weakens as the lattice becomes cubic, a structural-stability check near $y=0.2$ would test whether the proposed composition can actually form ordered L10.
  • The mechanism points toward a neighbouring scan over heavy lanthanides such as Gd or Dy, which could keep a large orbital moment while retaining more magnetization at room temperature; the paper does not perform that scan.
  • The paper does not report how many KKR-CPA runs the loop executed; publishing that count would let other groups quantify the speedup over the full 200,000-configuration space.
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

3 major / 6 minor

Summary. The paper presents a simulation-based autonomous search over a quaternary L10-FePt alloy space, combining KKR-CPA density functional theory with Bayesian optimization and an autoencoder. The search space, defined as Fe1-xXxPt1-yYy with x,y in steps of 0.01 up to 0.2, contains roughly 200,000 compositions. After a continuous 100-day run, the system frequently proposed Fe1-xMnxPt1-yEry, and a comprehensive KKR-CPA study of this system is reported. The authors claim that adding Mn and Er increases both the total magnetic moment M and the magnetocrystalline anisotropy energy EMCA relative to FePt, and they attribute the increases to the large Er local and orbital moments at zero temperature. The paper explicitly acknowledges that the result is a speculative prediction and that the L10 stability is unverified.

Significance. If the central claims hold, the paper would demonstrate a useful autonomous-search workflow for a large quaternary alloy space, with the merit that the final M and EMCA values come from direct ab initio calculations rather than from the machine-learning surrogate, so the conclusion is not circular. The paper also provides detailed composition-dependent local spin and orbital moments that give physical insight into the proposed mechanism. However, the practical significance as a recording-material discovery is currently conditional on finite-temperature behavior and structural stability, and the quantitative claims lack convergence tests and experimental benchmarks. With those additions, the work would constitute a solid contribution to computational materials discovery for magnetic alloys.

major comments (3)
  1. [Section 3, paragraph after Fig. 5(d)] The central practical claim that Fe1-xMnxPt1-yEry has high M and EMCA rests on zero-temperature KKR-CPA results: the large Er local moment (Fig. 5(d)) and orbital moment (Fig. 7(d)) are the stated origins of the increases in M and EMCA (Fig. 3). The authors themselves note that the magnetic moment of lanthanides such as Er 'decreases significantly at room temperature' and that the magnetization increase along the Er axis 'may not be observed' at 300 K. Because magnetic recording media operate near room temperature, this caveat means the reported high M and EMCA do not yet establish the promised practical advantage. The paper should either compute finite-temperature magnetization and magnetocrystalline anisotropy (e.g., via disordered local moment or mean-field approaches) or explicitly restrict the claim to zero-temperature predictions.
  2. [Section 2 (Methods) and Figure 3] No convergence tests or numerical error estimates are presented for the KKR-CPA calculations, and the calculated M and EMCA are not benchmarked against known values for pure L10-FePt. As the central result is a quantitative claim of 'superior' M and EMCA values (Fig. 3), the absence of any k-point/smearing/concentration-grid convergence checks, or a comparison with experimental FePt data, leaves the magnitude of the reported improvements unsupported. Please add convergence tests and a benchmark calculation of FePt M and EMCA against literature values, and state the numerical uncertainty appropriate to the 0.01 composition grid.
  3. [Section 3, final paragraph and Figure 4] The paper acknowledges that the stability of the L10 structures formed by Fe1-xMnxPt1-yEry is uncertain, and Figure 4 shows that the c/a ratio approaches unity for y = 0.2. Since all M and EMCA calculations assume a tetragonal L10 lattice, a cubic or non-L10 ground state would invalidate the proposed compositions as recording media. Formation enthalpies relative to competing phases, or at least a check of the assumed structure's stability, are needed to support the claim that the search identified 'new L10-based quaternary alloys.'
minor comments (6)
  1. [Section 1 and Figure 1 caption] The phrase 'magneto crystalline' should be 'magnetocrystalline' where it appears in the introduction and in the Figure 1 caption.
  2. [Section 3] The sentence 'lower M and EMCA compared than the initial data' should read 'compared with the initial data.'
  3. [Figures 2 and 3] The units for M and EMCA are not stated in the main text or on the figure axes; please specify them explicitly, for example as μB per formula unit and meV per formula unit.
  4. [Figure 2] The figure would benefit from clear axis labels and a legend identifying the initial and explored points, as the caption text alone is insufficient to distinguish the white and black circles.
  5. [References to Supplementary Materials] The manuscript refers to Supplementary Materials S1–S3 for calculation details; please ensure these are available to reviewers and that the main text summarizes the key settings such as exchange-correlation functional, k-point sampling, and treatment of the lattice parameter.
  6. [Equation (3)] The list of Y elements spans multiple lines; stating the total number of elements (38) explicitly would make the 'approximately 200,000 configurations' estimate transparent and reproducible.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the final FeMnPtEr property claim rests on direct KKR-CPA calculations, not on the ML surrogate or on author self-citations.

full rationale

The derivation chain is not circular. The machine-learning components (Bayesian optimization, autoencoder, Magpie descriptors) are used only to choose candidate compositions for the next ab initio calculation; every reported M and EMCA value, including the final Fe1-xMnxPt1-yEry maps in Figs. 3(a) and 3(b), is obtained by direct KKR-CPA evaluation rather than by reading values out of the fitted surrogate. The paper explicitly states that the ab initio phase 'computes M and EMCA based on the compositions recommended by the machine learning phase,' so the final claim of high M and EMCA is not equivalent to the model's training data or to the acquisition function. Citations to the authors' earlier works [19,21,22] are used to justify the method/descriptor design, but they are not load-bearing for the target result, do not invoke a uniqueness theorem, and do not forbid alternative compositions. The manuscript's own caveats about the room-temperature reduction of lanthanide moments and the unverified L10 stability are validity and robustness limitations, not circularity: they concern whether the zero-temperature computational result transfers to operating conditions, not whether the result reduces to its inputs. No fitted parameter is renamed as a prediction, no target quantity is defined in terms of itself, and no known result is repackaged under new coordinates. Therefore the paper's central computational claim is self-contained relative to its stated ab initio method, and no significant circularity is present.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No free parameters are fitted to the target result: the M and EMCA values come directly from KKR-CPA, and the ML surrogate is only used for candidate selection. The key axioms are the accuracy of KKR-CPA for disordered 3d-4f alloys, the assumed L10 structure, and the validity of zero-temperature magnetic moments at room temperature. No new physical entities are introduced.

assumptions (4)
  • domain assumption The KKR-CPA method (AkaiKKR) accurately computes total magnetic moment and magnetocrystalline anisotropy energy for disordered multicomponent alloys.
    The paper relies on this standard alloy DFT method to produce all M and EMCA values; no validation against experiment is provided (Methods, Section 2).
  • domain assumption The assumed L10 crystal structure with Fe/Mn on one sublattice and Pt/Er on the other is the relevant structure for the proposed alloys.
    The material space is defined as L10-based via Eq. (1), and the authors note that "the stability of the L10 structures formed by the elements in these alloys is uncertain" (Section 3).
  • domain assumption Magnetic properties computed at absolute zero are representative of the magnetic recording operating temperature.
    The paper explicitly flags that Er's moment decreases significantly at room temperature, implying the zero-temperature assumption is load-bearing (Section 3, after Fig. 5(d)).
  • domain assumption Composition increments of 0.01 and the CPA disorder model sufficiently sample the quaternary configurational space.
    The 200,000 compositions are defined by x,y in {0,...,0.2} in steps of 0.01 (Eq. 4), and only that discrete grid is searched (Methods, Section 2).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Autonomous materials search using machine learning and ab initio calculations for L10-FePt-based quaternary alloys." pith.science (2026). https://pith.science/paper/2E344BFL

@misc{pith2026241118907,
  author       = {Pith},
  title        = {Pith review of: Autonomous materials search using machine learning and ab initio calculations for L10-FePt-based quaternary alloys},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2E344BFL}},
  note         = {Machine review of arXiv:2411.18907}
}
read the original abstract

The efficient exploration of expansive material spaces remains a significant challenge in materials science. To address this issue, autonomous material search methods that combine machine learning with ab initio calculations have emerged as a promising solution. These approaches offer a systematic and rapid means of discovering new materials, particularly when the material space is too large. This requirement is particularly important in the development of L10-structured alloys as magnetic recording media. These materials require a high magnetic moment (M) and magnetocrystalline anisotropy energy (EMCA) to satisfy the demands of next-generation data storage technologies. Although autonomous search methods have been successfully applied to various material systems, quaternary L10 alloys with optimized magnetic properties remain an open and underexplored frontier. In this study, we present a simulation-based autonomous search method aimed at identifying quaternary L10 alloys with enhanced M and EMCA values. Over a continuous 100-day search, our system suggested the FeMnPtEr alloy system as a promising candidate, exhibiting superior values for both M and EMCA. Although further experimental validation is required, this study underscores the potential of autonomous search methods to accelerate the discovery of advanced materials. Keywords: L10, FePt, machine learning, ab initio calculations, Bayesian optimization

Figures

Figures reproduced from arXiv: 2411.18907 by the authors.

Figure 1
Figure 1. It includes sequential ab initio calculations and machine learning. The ab initio phase calculates M and EMCA based on the compositions recommended by the machine learning phase. The data accumulated from these calculations guided the composition choices for subsequent ab initio calculations. The ab initio calculations used Green’s function-based density functional theory using the Korringa–Kohn–Rostoker coherent po… view at source ↗
Figure 1
Figure 1. Overview of an autonomous material search system [PITH_FULL_IMAGE:figures/full_fig_p015_1.png] view at source ↗
Figure 2
Figure 2. Results of the autonomous search for materials with high [PITH_FULL_IMAGE:figures/full_fig_p016_2.png] view at source ↗
Figures from the paper (1 more)
Figure 3
Figure 3. Figure 3: Composition dependence of the (a) total magnetic moment (M) and (b) magnetocrystalline anisotropy energy (EMCA) in the Fe1-xMnxPt1-yEry system [PITH_FULL_IMAGE:figures/full_fig_p017_3.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 21 canonical work pages

  1. [1]

    A robotic platform for flow synthesis of organic compounds informed by AI planning

    Coley CW, Thomas DA, Lummiss JA, et al. A robotic platform for flow synthesis of organic compounds informed by AI planning. Science. 2019;365(6453):eaax1566. doi: 10.1126/science.aax1566

    doi:10.1126/science.aax1566 2019

  2. [2]

    A mobile robotic chemist

    Burger B, Maffettone PM, Gusev VV, et al. A mobile robotic chemist. Nature. 2020;583(7815):237–241. doi: 10.1038/s41586-020-2442-2

    doi:10.1038/s41586-020-2442-2 2020

  3. [3]

    Autonomy in materials research: a case study in carbon nanotube growth

    Nikolaev P, Hooper D, Webber F, et al. Autonomy in materials research: a case study in carbon nanotube growth. npj Comput Mater. 2016;2(1):16031. doi: 10.1038/npjcompumats.2016.31

    work page doi:10.1038/npjcompumats.2016.31 2016

  4. [4]

    Autonomous materials synthesis by machine learning and robotics

    Shimizu R, Kobayashi S, Watanabe Y, et al. Autonomous materials synthesis by machine learning and robotics. APL Mater. 2020;8(11):111110. doi: 10.1063/5.0020370

    doi:10.1063/5.0020370 2020

  5. [5]

    Robot-accelerated perovskite investigation and discovery

    Li Z, Najeeb MA, Alves L, et al. Robot-accelerated perovskite investigation and discovery. Chem Mater. 2020;32(13):5650–5663. doi: 10.1021/acs.chemmater.0c01153

    work page doi:10.1021/acs.chemmater.0c01153 2020

  6. [6]

    ChemOS: orchestrating autonomous experimentation

    Roch LM, Häse F, Kreisbeck C, et al. ChemOS: orchestrating autonomous experimentation. Sci Robot. 2018;3(19):eaat5559. doi: 10.1126/scirobotics.aat5559

    doi:10.1126/scirobotics.aat5559 2018

  7. [7]

    Closed-loop optimization of fast-charging protocols for batteries with machine learning

    Attia PM, Grover A, Jin N, et al. Closed-loop optimization of fast-charging protocols for batteries with machine learning. Nature. 2020;578(7795):397–402. doi: 10.1038/s41586-020-1994-5

    doi:10.1038/s41586-020-1994-5 2020

  8. [8]

    Controlling an organic synthesis robot with machine learning to search for new reactivity

    Granda JM, Donina L, Dragone V, et al. Controlling an organic synthesis robot with machine learning to search for new reactivity. Nature. 2018;559(7714):377–381. doi: 10.1038/s41586-018-0307-8

    work page doi:10.1038/s41586-018-0307-8 2018

Show all 35 references
  1. [9]

    Toward autonomous design and synthesis of novel inorganic materials

    Szymanski NJ, Zeng Y, Huo H, et al. Toward autonomous design and synthesis of novel inorganic materials. Mater Horiz. 2021;8(8):2169–2198. doi: 10.1039/D1MH00495F

    2021 doi

  2. [10]

    Autonomous experiments using active learning and AI

    Ren Z, Ren Z, Zhang Z, et al. Autonomous experiments using active learning and AI. Nat Rev Mater. 2023;8(9):563–564. doi: 10.1038/s41578-023-00588-4

    2023 doi

  3. [11]

    An autonomous laboratory for the accelerated synthesis of novel materials

    Szymanski NJ, Rendy B, Fei Y, et al. An autonomous laboratory for the accelerated synthesis of novel materials. Nature. 2023;624(7990):86–91. doi: 10.1038/s41586-023-06734-w

    2023 doi

  4. [13]

    Machine learning autonomous identification of magnetic alloys beyond the Slater-Pauling limit

    Iwasaki Y, Sawada R, Saitoh E, et al. Machine learning autonomous identification of magnetic alloys beyond the Slater-Pauling limit. Commun Mater. 2021;2(1):31. doi: 10.1038/s43246-021-00135-0

    2021 doi

  5. [14]

    Boosting material modeling using game tree search

    Sawada R, Iwasaki Y, Ishida M. Boosting material modeling using game tree search. Phys Rev Mater. 2018;2(10):103802. doi: 10.1103/PhysRevMaterials.2.103802

    2018 doi

  6. [15]

    Prediction of low-thermal-conductivity compounds with first-principles anharmonic lattice-dynamics calculations and bayesian optimization

    Seko A, Togo A, Hayashi H, et al. Prediction of low-thermal-conductivity compounds with first-principles anharmonic lattice-dynamics calculations and bayesian optimization. Phys Rev Lett. 2015;115(20):205901. doi: 10.1103/PhysRevLett.115.205901

    2015 doi

  7. [16]

    Bayesian-driven first-principles calculations for accelerating exploration of fast ion conductors for rechargeable battery application

    Jalem R, Kanamori K, Takeuchi I, et al. Bayesian-driven first-principles calculations for accelerating exploration of fast ion conductors for rechargeable battery application. Sci Rep. 2018;8(1):5845. doi: 10.1038/s41598-018-23852-y

    2018 doi

  8. [17]

    Exploration of a large-scale reconstructed structure on GaN(0001) surface by Bayesian optimization

    Kusaba A, Kangawa Y, Kuboyama T, et al. Exploration of a large-scale reconstructed structure on GaN(0001) surface by Bayesian optimization. Appl Phys Lett. 2022;120(2):021602. doi: 10.1063/5.0078660

    2022 doi

  9. [18]

    Autonomous synthesis system integrating theoretical, informatics, and experimental approaches for large- magnetic-anisotropy materials

    Furuya D, Miyashita T, Miura Y, et al. Autonomous synthesis system integrating theoretical, informatics, and experimental approaches for large- magnetic-anisotropy materials. Sci Technol Adv Mater Meth. 2022;2(1):280–

    2022

  10. [19]

    Efficient autonomous material search method combining ab initio calculations, autoencoder, and multi-objective Bayesian optimization

    Iwasaki Y, Hwang J, Sakuraba Y, et al. Efficient autonomous material search method combining ab initio calculations, autoencoder, and multi-objective Bayesian optimization. Sci Technol Adv Mater. 2022;2(1):365–371. doi: 10.1080/27660400.2022.2123263

    2022

  11. [20]

    Improving efficiency of autonomous material search via transfer learning from nontarget properties

    Hwang J, Iwasaki Y. Improving efficiency of autonomous material search via transfer learning from nontarget properties. Sci Technol Adv Mater. 2023;3(1):2254202. doi: 10.1080/27660400.2023.2254202

    2023

  12. [22]

    Autonomous search for half-metallic materials with B2 structure

    Iwasaki Y, Toyama R, Yamazaki T, et al. Autonomous search for half-metallic materials with B2 structure. Sci Technol Adv Mater Meth. 2024;4(1):2403966. doi: 10.1080/27660400.2024.2403966

    2024

  13. [23]

    Heat Assisted Magnetic Recording

    Kryder MH, Gage EC, McDaniel TW, et al. Heat Assisted Magnetic Recording. Proc IEEE Inst Electr Electron Eng. 2008;96(11):1810-1835. doi: 10.1109/JPROC.2008.2004315

    2008

  14. [24]

    Determination of the anisotropy constant and saturation magnetization and magnetic properties of powers of an iron-platinum alloy

    Ivanov OA, Solina LV, Demshina VA, et al. Determination of the anisotropy constant and saturation magnetization and magnetic properties of powers of an iron-platinum alloy. Phys Met Metallog. 1973;35:81

    1973

  15. [25]

    Thermal effect limits in ultrahigh-density magnetic recording

    Weller D, Moser A. Thermal effect limits in ultrahigh-density magnetic recording. IEEE Trans Magn. 1999;35(6):4423-4439. doi: 10.1109/20.809134

    1999 doi

  16. [26]

    High Ku materials approach to 100 Gbits/in2

    Weller D, Moser A, Folks L, et al. High Ku materials approach to 100 Gbits/in2. IEEE Trans Magn. 2000;36:10-15

    2000

  17. [27]

    Ab-initio electronic-structure calculation code

    Akai KKR (machikaneyama). Ab-initio electronic-structure calculation code. Available from: http://kkr.issp.u-tokyo.ac.jp

  18. [28]

    Electronic structure Ni-Pd alloys calculated by the self-consistent KKR- CPA method

    Akai H. Electronic structure Ni-Pd alloys calculated by the self-consistent KKR- CPA method. J Phys Soc Jpn. 1982;51(2):468–474. doi: 10.1143/JPSJ.51.468

    1982 doi

  19. [29]

    Statistical physics of multicomponent alloys using KKR-CPA

    Khan SN, Staunton JB, Stocks GM. Statistical physics of multicomponent alloys using KKR-CPA. Phys Rev B. 2016;93(5):054206. doi: 10.1103/PhysRevB.93.054206

    2016 doi

  20. [30]

    Investigation of the site preference in Mn2RuSn using KKR-CPA-LDA calculation

    Yang L, Liu B, Luo H, et al. Investigation of the site preference in Mn2RuSn using KKR-CPA-LDA calculation. J Magn Mater. 2015;382(15):247–251. doi: 10.1016/j.jmmm.2015.01.081

    2015 doi

  21. [31]

    Fast Korringa-Kohn-Rostoker coherent potential approximation and its application to FCC Ni-Fe systems

    Akai H. Fast Korringa-Kohn-Rostoker coherent potential approximation and its application to FCC Ni-Fe systems. J Phys Condens Matter. 1989;1(43):8045–

    1989

  22. [32]

    Practical bayesian optimization of machine learning algorithms

    Snoek J, Larochelle H, Adams RP. Practical bayesian optimization of machine learning algorithms. Adv Neural Inf Process Syst. 2012;25:2960–2968

    2012

  23. [33]

    Reducing the dimensionality of data with neural networks

    Hinton GE, Salakhutdinov RR. Reducing the dimensionality of data with neural networks. Science. 2006;313(5786):504–507. doi: 10.1126/science.1127647

    2006 doi

  24. [34]

    A general-purpose machine learning framework for predicting properties of inorganic materials

    Ward L, Agrawal A, Choudhary A, et al. A general-purpose machine learning framework for predicting properties of inorganic materials. npj Comput Mater. 2016;2(1):16028. doi: 10.1038/npjcompumats.2016.28

    2016 doi

  25. [35]

    Using confidence bounds for exploitation-exploration trade-off

    Auer P. Using confidence bounds for exploitation-exploration trade-off. J Mach Learn Res. 2002;3:397–422

    2002

  26. [36]

    Tight-binding approach to the orbital magnetic moment and magnetocrystalline anisotropy of transition-metal monolayers

    Bruno P. Tight-binding approach to the orbital magnetic moment and magnetocrystalline anisotropy of transition-metal monolayers. Phys Rev B. 1989;39(1):865-868. doi: 10.1103/PhysRevB.39.865 Figures and captions Figure 1. Overview of an autonomous material search system; the ab...

    1989 doi

  27. [8064]

    doi: 10.1088/0953-8984/1/43/006

    doi

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.