Improved hardness prediction for reduced-activation high-entropy alloys by incorporating symbolic regression and domain adaptation on small datasets

Abstract

The reduced-activation high-entropy alloys (RAHEAs) have promising applications in advanced nuclear systems due to their low activation, excellent mechanical properties and radiation resistance. However, compared to the conventional high-entropy alloys (HEAs), the relatively small datasets of RAHEAs pose challenges for alloy design by using conventional machine learning (ML) methods. In this work, we proposed a framework by incorporating symbolic regression (SR) and domain adaptation to improve the accuracy of property prediction based on the small datasets of RAHEAs. The conventional HEA datasets and RAHEA datasets were classified as source and target domains, respectively. SR was used to generate features from element-based features in the source domains. The domain-invariant features related to hardness were captured and used to construct the ML model, which significantly improved the prediction accuracy for both HEAs and RAHEAs. The normalized root mean square error decreases by 24% for HEAs and 30% for RAHEAs compared to that of the models trained with element-based features. The proposed framework can achieve accurate and robust prediction on small datasets with interpretable domain-invariant features. This research paves the way for efficient material design under small dataset scenarios.

Keywords

High-entropy alloy / machine learning / small dataset / symbolic regression / domain adaptation

References

[1] Yeh J,Lin S.Nanostructured high-entropy alloys with multiple principal elements: novel alloy design concepts and outcomes.Adv Eng Mater2004;6:299-303

[2] Li Z,Deng Y,Tasan CC.Metastable high-entropy dual-phase alloys overcome the strength-ductility trade-off.Nature2016;534:227-30

[3] Lei Z,Wu Y.Enhanced strength and ductility in a high-entropy alloy via ordered oxygen complexes.Nature2018;563:546-50

[4] Wang D,Li C,Guo S.Enhanced creep resistance of Ti₃₀Al₂₅Zr₂₅Nb₂₀ high-entropy alloy at room temperature.J Alloys Compd2021;885:161038

[5] Su Z,Yang J.The effect of interstitial carbon atoms on defect evolution in high entropy alloys under helium irradiation.Acta Mater2022;233:117955

[6] Yeh J.Physical metallurgy of high-entropy alloys.JOM2015;67:2254-61

[7] Islam N,Zhuang HL.Machine learning for phase selection in multi-principal element alloys.Comput Mater Sci2018;150:230-5

[8] Wen C,Wang C.Machine learning assisted design of high entropy alloys with desired property.Acta Mater2019;170:109-17

[9] Xiong J,Zhang T.Machine learning of phases and mechanical properties in complex concentrated alloys.J Mater Sci Technol2021;87:133-42

[10] Wen C,Zhang Y.Modeling solid solution strengthening in high entropy alloys using machine learning.Acta Mater2021;212:116917

[11] Vazquez G,Sauceda D.Efficient machine-learning model for fast assessment of elastic properties of high-entropy alloys.Acta Mater2022;232:117924

[12] Li X,Li C,Ding W.Accelerated design of low-activation high entropy alloys with desired phase and property by machine learning.Appl Mater Today2024;36:102000

[13] Miracle D.A critical review of high entropy alloys and related concepts.Acta Mater2017;122:448-511

[14] Gorsse S,Senkov ON.Database on the mechanical properties of high entropy alloys and complex concentrated alloys.Data Brief2018;21:2664-78 PMCID:PMC6290247

[15] Xue D,Yuan R.An informatics approach to transformation temperatures of NiTi-based shape memory alloys.Acta Mater2017;125:532-41

[16] Jones R,Mccarthy K.Low activation materials.J Nucl Mater1999;271-272:518-25

[17] Xian X,Zhang B.A high-entropy V₃₅Ti₃₅Fe₁₅Cr₁₀Zr₅ alloy with excellent high-temperature strength.Mater Design2017;121:229-36

[18] Barron P,Fellowes J,Dawson H.Towards V-based high-entropy alloys for nuclear fusion applications.Scr Mater2020;176:12-6

[19] Kareer A,Li B,Armstrong D.Short communication: ‘Low activation, refractory, high entropy alloys for nuclear applications’.J Nucl Mater2019;526:151744

[20] Wang Y,Kwok JT.Generalizing from a few examples: a survey on few-shot learning.ACM Comput Surv2021;53:1-34

[21] Zhang Y.A strategy to apply machine learning to small datasets in materials science.npj Comput Mater2018;4:81

[22] Pan SJ.A survey on transfer learning.IEEE Trans Knowl Data Eng2010;22:1345-59

[23] Zhuang F,Duan K.A comprehensive survey on transfer learning.Proc IEEE2021;109:43-76

[24] Kim Y,Yang C,Gu GX.Deep learning framework for material design space exploration using active transfer learning and data augmentation.npj Comput Mater2021;7:609

[25] Wang X,Zeng S,Chen Y.Element-wise representations with ECNet for material property prediction and applications in high-entropy alloys.npj Comput Mater2022;8:945

[26] Wei X,Jia Z.High-cycle fatigue S-N curve prediction of steels based on a transfer learning-guided convolutional neural network.J Mater Inf2022;2:9

[27] Quiñonero-Candela J,Schwaighofer A.Dataset shift in machine learning. MIT Press, 2009.

[28] Li K,Choudhary K,Hattrick-Simpers J.A critical examination of robustness and generalizability of machine learning prediction of materials properties.npj Comput Mater2023;9:1012

[29] Wang J,Liu C.Generalizing to unseen domains: a survey on domain generalization.IEEE Trans Knowl Data Eng2023;35:8052-72

[30] Motiian S,Iranmanesh SM.Few-shot adversarial domain adaptation. In Proceedings of the 31st International Conference on Neural Information Processing Systems, Long Beach, USA; Curran Associates Inc., 2017; pp 6673-83.

[31] Long M,Ding G,Yu PS.Adaptation regularization: a general framework for transfer learning.IEEE Trans Knowl Data Eng2014;26:1076-89

[32] Stojanov P,Gong M,Carbonell JG.Domain adaptation with invariant representation learning: what transformations to learn? In Proceedings of the 35th International Conference on Neural Information Processing Systems; Curran Associates Inc., 2021.

[33] Goetz A,Müller M.Addressing materials’ microstructure diversity using transfer learning.npj Comput Mater2022;8:703

[34] Hu J,Fu N.Realistic material property prediction using domain adaptation based machine learning.Digit Discov2024;3:300-12

[35] Zhao H,Zhang K. On learning invariant representations for domain adaptation. arXiv 2019, arXiv:1901.09453. Available online: https://doi.org/10.48550/arXiv.1901.09453. (accessed 6 Feb 2025)

[36] Huo H.Unified representation of molecules and crystals for machine learning.Mach Learn Sci Technol2022;3:045017

[37] Pan SJ,Kwok JT.Domain adaptation via transfer component analysis.IEEE Trans Neural Netw2011;22:199-210

[38] Fernando B,Sebban M.Unsupervised visual domain adaptation using subspace alignment. In 2013 IEEE International Conference on Computer Vision, Sydney, Australia, December 01-08, 2013; IEEE, 2013; pp 2960-7.

[39] Peng H,Ding C.Feature selection based on mutual information: criteria of max-dependency, max-relevance, and min-redundancy.IEEE Trans Pattern Anal Mach Intell2005;27:1226-38

[40] Koza J.Genetic programming as a means for programming computers by natural selection.Stat Comput1994;4:BF00175355

[41] Chen Y,Zhao J,Li J.Symbolic regression and two-point statistics assisted structure-property linkage based on irregular-representative volume element.Adv Theory Simul2023;6:2200524

[42] Xiong J,Zhang T.A machine-learning approach to predicting and understanding the properties of amorphous metallic alloys.Mater Design2020;187:108378

[43] Huang X,Xu H.Predicting and understanding the ductility of BCC high entropy alloys via knowledge-integrated machine learning.Mater Design2024;239:112797

[44] Lee JA,Park H,Kim HS.Unveiling yield strength of metallic materials using physics-enhanced machine learning under diverse experimental conditions.Acta Mater2024;275:120046

[45] Feng S,Chen G,Chen X.LBE corrosion fatigue life prediction of T91 steel and 316 SS using machine learning method assisted by symbol regression.Int J Fatigue2023;177:107962

[46] Couzinié JP,Miracle DB.Comprehensive data compilation on the mechanical properties of refractory high-entropy alloys.Data Brief2018;21:1622-41

[47] Carruthers A,Rigby M.Novel reduced-activation TiVCrFe based high entropy alloys.J Alloys Compd2021;856:157399

[48] Ayyagari A,Muskeri S.Low activation high entropy alloys for next generation nuclear applications.Materialia2018;4:99-103

[49] Liu Y,Zhang P.Enhanced strength without sacrificing ductility in FeCrMnVSi_x high entropy alloys via controlling the ratio of metallic to covalent bonding.Mater Design2023;225:111565

[50] Carruthers A,Hardwick L,Gandy A.An assessment of the high-entropy alloy system VCrMnFeAl_x.J Alloys Compd2021;888:161525

[51] Haynes WM.CRC handbook of chemistry and physics. 97th edition. CRC Press, 2016.

[52] Guo S.Phase stability in high entropy alloys: formation of solid-solution phase or amorphous phase.Prog Nat Sci Mater Int2011;21:433-46

[53] Wang WY,Wang Y.Atomic and electronic basis for the serrations of refractory high-entropy alloys.npj Comput Mater2017;3:24

[54] Weng B,Zhu R.Simple descriptor derived from symbolic regression accelerating the discovery of new perovskite catalysts.Nat Commun2020;11:3513 PMCID:PMC7360597

[55] Lavalle SM,Lindemann SR.On the relationship between classical grid search and probabilistic roadmaps.Int J Robot Res2004;23:673-92

[56] Lundberg SM.A unified approach to interpreting model predictions. In Proceedings of the 31st International Conference on Neural Information Processing Systems, Long Beach, USA; Curran Associates Inc., 2017; pp 4766-75.

[57] Pedregosa F,Gramfort A. Scikit-learn: machine learning in Python. arXiv 2012, arXiv:1201.0490. Available online: https://doi.org/10.48550/arXiv.1201.0490. (accessed 6 Feb 2025)

[58] Harris CR,van der Walt SJ.Array programming with NumPy.Nature2020;585:357-62 PMCID:PMC7759461

[59] Virtanen, P.; Gommers, R.; Oliphant, T. E.; et al; SciPy 1.0 Contributors. SciPy 1.0: fundamental algorithms for scientific computing in Python. Nat. Methods. 2020, 17, 261-72. PMCID:PMC7056644

[60] McKinney W.Data structures for statistical computing in Python. 2010.

[61] Borgwardt KM,Rasch MJ,Schölkopf B.Integrating structured biological data by kernel maximum mean discrepancy.Bioinformatics2006;22:e49-57

[62] Du YY,Li TJ,Zhang GL.Effect of aluminium content of Al_xCrFe_1.5Ni_0.5 multiprincipal alloys on microstructure and alloy hardness.Mater Res Innov2011;15:107-10

[63] Wang W,Wang S,Lai C.Effects of Al addition on the microstructure and mechanical property of Al_xCoCrFeNi high-entropy alloys.Intermetallics2012;26:44-51

[64] Wilcoxon F.Individual comparisons by ranking methods.Biometrics Bull1945;1:80