有监督学习算法在材料科学中的应用

doi:10.11871/jfdc.issn.2096-742X.2023.04.004

数据与计算发展前沿 ›› 2023, Vol. 5 ›› Issue (4): 38-47.

CSTR: 32002.14.jfdc.CN10-1649/TP.2023.04.004

doi: 10.11871/jfdc.issn.2096-742X.2023.04.004

• 专刊: “基础研究”联合专刊 • 上一篇下一篇

有监督学习算法在材料科学中的应用

刘端阳¹(),魏钟鸣^1,^2,^*()

1.中国科学院半导体研究所，超晶格国家重点实验室, 北京 100083
2.中国科学院大学，材料科学与光电技术学院, 北京 100049

收稿日期:2023-06-30 出版日期:2023-08-20 发布日期:2023-08-23
通讯作者: *魏钟鸣（E-mail: zmwei@semi.ac.cn）
作者简介:刘端阳，中国科学院半导体研究所超晶格国家重点实验室，副研究员，从事半导体材料的第一性原理计算研究，主要研究方向为机器学习在计算物理中的应用。
本文负责素材收集和论文初稿撰写。
LIU Duanyang is an associate researcher at the State Key Laboratory of Superlattices in the Institute of Semiconductors, Chinese Academy of Sciences. He is engaged in first-principles calculations of semiconductor materials and his main research focus is on the application of machine learning in computational physics.
He is responsible for material collection and the initial drafting of research papers.
E-mail: liudy@semi.ac.cn|魏钟鸣，中国科学院半导体研究所超晶格国家重点实验室，研究员，博士生导师，国家杰出青年科学基金获得者，在国内外知名杂志上发表论文50余篇，合作出版专著1部。授权发明专利12项，其中5项已经完成技术转让。主要研究方向为低维半导体材料与光电器件的实验和理论研究。
本文负责制定论文框架与稿件修订工作。
WEI Zhongming is a researcher and doctoral supervisor at the State Key Laboratory of Superlattices in the Institute of Semiconductors, Chinese Academy of Sciences. He is a recipient of the National Science Fund for Distinguished Young Scholars and has published over 50 papers in renowned domestic and international journals.
He has also co-authored one monograph and holds 12 granted invention patents, with 5 of them already transferred for technology commercialization. His main research focus is on the experimental and theoretical study of low-dimensional semiconductor materials and optoelectronic devices.
He is responsible for formulating the framework of research papers and revising the manuscripts.
E-mail: zmwei@semi.ac.cn
基金资助:
中国科学院网络安全和信息化专项应用示范培育项目“集成电路用单晶硅加工工艺的人工智能辅助软件与平台”(CAS-WX2023PY-0101)

Application of Supervised Learning Algorithms in Materials Science

LIU Duanyang¹(),WEI Zhongming^1,^2,^*()

1. State Key Laboratory of Superlattices and Microstructures, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China
2. Center of Materials Science and Optoelectronics Engineering, University of Chinese Academy of Sciences, Beijing 100049, China

Received:2023-06-30 Online:2023-08-20 Published:2023-08-23

摘要/Abstract

摘要：

【目的】本文希望对近年来机器学习在材料学研究中的应用做一概略的介绍，为相关的研究提供一定的参考。【文献范围】本文主要参考引述了近几年来材料数据库相关文献，以及使用机器学习算法进行材料性能预测、发现新材料的研究论文。【方法】本文介绍了有监督机器学习的处理流程，并介绍了多种有监督机器学习算法在材料科学中的应用现状。【结果】机器学习算法，帮助总结了材料性能与材料的组成元素、晶格结构等的规律，对发现新材料具有重要的意义，而机器学习力场方法则展现出处理复杂的相变、界面等问题的潜力。【局限】鉴于目前掌握的研究水平，主要重点介绍的是有监督机器学习方法在材料性能预测等几个领域的应用，对于无监督学习以及其他材料研究领域的引述尚缺乏。【结论】这是一个新兴的领域，未来将成为材料科学的一个重要组成部分。

关键词: 机器学习, 材料科学, 神经网络, 算法, 性能预测

Abstract:

[Objective] This article aims to provide a brief overview of the applications of machine learning in materials research in recent years, offering a reference for related studies. [Literature Scope] Therefore, this article mainly references recent literature and materials databases and research papers utilizing machine learning algorithms for material property prediction and new material discovery. [Methods] The article introduces the workflow of supervised machine learning and presents the current applications of various supervised machine learning algorithms in materials science. [Results] Machine learning algorithms help to identify patterns between material properties and factors such as composition elements and crystal structures, making them significant in the discovery of new materials. Additionally, force field methods using machine learning demonstrate potential in addressing complex phenomena like phase transitions and interfaces. [Limitations] Due to the limitations of the author’s expertise, the focus of the article is primarily on the application of supervised machine learning methods in material property prediction and a few other areas. Citations regarding to unsupervised learning and other research fields in materials science are currently inadequate. [Conclusions] This is an emerging field that is expected to become an important component of materials science in the future.

Key words: machine learning, materials science, neural networks, algorithms, properties prediction

刘端阳, 魏钟鸣. 有监督学习算法在材料科学中的应用[J]. 数据与计算发展前沿, 2023, 5(4): 38-47.

LIU Duanyang, WEI Zhongming. Application of Supervised Learning Algorithms in Materials Science[J]. Frontiers of Data and Computing, 2023, 5(4): 38-47, https://cstr.cn/32002.14.jfdc.CN10-1649/TP.2023.04.004.

图/表 4

参考文献 52

[1]	JAIN A, ONG S P, HAUTIER G, et al. Commentary: The Materials Project: A materials genome approach to accelerating materials innovation[J]. APL materials, 2013, 1(1): 011002. doi: 10.1063/1.4812323
[2]	HAASTRUP S, STRANGE M, PANDEY M, et al. The Computational 2D Materials Database: high-throughput modeling and discovery of atomically thin crystals[J]. 2D Materials, 2018, 5(4): 042002. doi: 10.1088/2053-1583/aacfc1
[3]	CURTAROLO S, SETYAWAN W, HART G L W, et al. AFLOW: an automatic framework for high-throughput materials discovery[J]. Computational Material Science, 2012, 58: 218-226. doi: 10.1016/j.commatsci.2012.02.005
[4]	CURTAROLO S, SETYAWAN W, WANG S, et al. AFLO-WLIB. ORG: A distributed materials properties repo-sitory from high-throughput ab initio calculations[J]. Comp-utational Materials Science, 2012, 58: 227-235.
[5]	BLUM L C, REYMOND J L. 970 million druglike small mol-ecules for virtual screening in the chemical universe data-base GDB-13[J]. Journal of the American Chemical Soc-iety, 2009, 131(25): 8732-8733.
[6]	RAMAKRISHNAN R, DRAL P O, RUPP M, et al. Qu-antum chemistry structures and properties of 134 kilo molecules[J]. Scientific data, 2014, 1(1): 1-7.
[7]	TAO Q, LU T, SHENG Y, et al. Machine learning aided design of perovskite oxide materials for photocatalytic water splitting[J]. Journal of Energy Chemistry, 2021, 60: 351-359. doi: 10.1016/j.jechem.2021.01.035
[8]	XIE T, GROSSMAN J C. Crystal graph convolutional neural networks for an accurate and interpretable predi-ction of material properties[J]. Physical review letters, 2018, 120(14): 145301. doi: 10.1103/PhysRevLett.120.145301
[9]	X Y, LYU H Y, HAO K R, et al. Large family of two-dimensional ferroelectric metals discovered via machine learning[J]. Science Bulletin, 2021, 66(3): 233-242. doi: 10.1016/j.scib.2020.09.010
[10]	ZHANG Y, LING C. A strategy to apply machine learn-ing to small datasets in materials science[J]. Npj Compu-tational Materials, 2018, 4(1): 25.
[11]	KHMAISSIA F, FRIGUI H, ANDRIOTIS A N, et al. Data driven modeling of magnetism in dilute magnetic semiconductors: correlation between the magnetic featur-es of diluted magnetic semiconductors and electronic properties of the constituent atoms[J]. Journal of Physics: Condensed Matter, 2019, 31(44): 445901. doi: 10.1088/1361-648X/ab31d6
[12]	HANSEN K, MONTAVON G, BIEGLER F, et al. Ass-essment and validation of machine learning methods for predicting molecular atomization energies[J]. Journal of Chemical Theory and Computation, 2013, 9(8): 3404-3419. doi: 10.1021/ct400195d
[13]	HOU Z, TAKAGIWA Y, SHINOHARA Y, et al. Machine-learning-assisted development and theoretical conside-ration for the Al2Fe3Si3 thermoelectric material[J]. ACS applied materials & interfaces, 2019, 11(12): 11545-11554.
[14]	JINNOUCHI R, LAHNSTEINER J, KARSAI F, et al. Phase transitions of hybrid perovskites simulated by machine-lea-rning force fields trained on the fly with Bayesian infere-nce[J]. Physical review letters, 2019, 122(22): 225701. doi: 10.1103/PhysRevLett.122.225701
[15]	HU L, HUANG B, LIU F. Atomistic Mechanism Unde-rlying the Si (111)-(7×7) Surface Reconstruction Revea-led by Artificial Neural-Network Potential[J]. Physical Review Letters, 2021, 126(17): 176101. doi: 10.1103/PhysRevLett.126.176101
[16]	ORTES C, VAPNIK V. Support vector network[J]. Mach-ine Learning, 1995, 20: 1-25.
[17]	BOSER B E, GUYON I M, VAPNIK V N. A training algorithm for optimal margin classifiers[C]// Proceedings of the fifth annual workshop on Computational learning theory. 1992: 144-152.
[18]	OLATUNJI S O, OWOLABI T O. Barium titanate semic-onductor band gap characterization through gravitatio-nally optimized support vector regression and extreme learning machine computational methods[J]. Mathema-tical Problems in Engineering, 2021, 2021: 1-12.
[19]	HUANG Y, YU C, CHEN W, et al. Band gap and band align-ment prediction of nitride-based semiconductors using machine learning[J]. Journal of Materials Che-mistry C, 2019, 7(11): 3238-3245.
[20]	SHAMSAH S M I, OWOLABI T O. Newtonian mec-hanics based hybrid machine learning method of charact-erizing energy band gap of doped zno semico-nductor[J]. Chinese Journal of Physics, 2020, 68: 493-506. doi: 10.1016/j.cjph.2020.10.002
[21]	WESTON L, STAMPFL C. Machine learning the band gap properties of kesterite I2-II-IV-V4 quaternary com-pounds for photovoltaics applications[J]. Physical Rev-iew Materials, 2018, 2(8): 085407.
[22]	TAWFIK S A, ISAYEV O, SPENCER M J S, et al. Pred-icting thermal properties of crystals using machine learn-ing[J]. Advanced Theory and Simulations, 2020, 3 (2): 1900208. doi: 10.1002/adts.v3.2
[23]	TONG W, WEI Q, YAN H Y, et al. Accelerating inverse crystal structure prediction by machine learning: A case study of carbon allotropes[J]. Frontiers of Physics, 2020, 15: 1-7.
[24]	CRISTIANINI N, SHAWE-TAYLOR J. An introduction to support vector machines and other kernel-based lear-ning methods[M]. Cambridge: Cambridge university press, 2000.
[25]	MILARDOVICH D, JECH M, WALDHOER D, et al. Machine Learning Prediction of Defect Structures in Amorphous Silicon Dioxide[C]// ESSDERC 2021-IEEE 51st European Solid-State Device Research Conference (ESSDERC). IEEE, 2021: 239-242.
[26]	KUMAR U, NAYAK S, CHAKRABARTY S, et al. Gall-ium-Boron-Phosphide (GaBP 2): a new III-V semicon-ductor for photovoltaics[J]. Journal of Materials Science, 2020, 55(22): 9448-9460. doi: 10.1007/s10853-020-04631-5
[27]	POLAK M P, JACOBS R, MANNODI-KANAKKIT-HODI A, et al. Machine learning for impurity charge-state transition levels in semiconductors from elemental properties using multi-fidelity datasets[J]. The Journal of Chemical Physics, 2022, 156(11): 114110. doi: 10.1063/5.0083877
[28]	ZHANG Y, LING C. A strategy to apply machine lea-rning to small datasets in materials science[J]. Npj Com-putational Materials, 2018, 4(1): 25.
[29]	ATAHAN-EVRENK S, ATALAY F B. Prediction of intramolecular reorganization energy using machine lear-ning[J]. The Journal of Physical Chemistry A, 2019, 123(36): 7855-7863. doi: 10.1021/acs.jpca.9b02733
[30]	LEDERER J, KAISER W, MATTONi A, et al. Machine learning—based charge transport computation for pentacene[J]. Advanced Theory and Simulations, 2019, 2(2): 1800136. doi: 10.1002/adts.v2.2
[31]	FABER F A, HUTCHISON L, HUANG B, et al. Pred-iction errors of molecular machine learning models lower than hybrid DFT error[J]. Journal of chemical theory and computation, 2017, 13(11): 5255-5264. doi: 10.1021/acs.jctc.7b00577
[32]	BREIMAN L. Classification and regression trees[M]. London: Routledge, 2017.
[33]	QUINLAN J R. Induction of decision trees[J]. Machine learning, 1986, 1: 81-106.
[34]	QUINLAN J R. C4. 5: programs for machine learn-ing[M]. Elsevier, 2014.
[35]	HO T K. Random decision forests[C]// Proceedings of 3rd international conference on document analysis and reco-gnition. IEEE, 1995, 1: 278-282.
[36]	ZHANG Y, XU W, LIU G, et al. Bandgap prediction of two-dimensional materials using machine learning[J]. PLOS ONE, 2021, 16(8): e0255637. doi: 10.1371/journal.pone.0255637
[37]	WANG T, ZHANG K, THÉ J, et al. Accurate prediction of band gap of materials using stacking machine learning model[J]. Computational Materials Science, 2022, 201: 110899. doi: 10.1016/j.commatsci.2021.110899
[38]	ZHU Z, DONG B, GUO H, et al. Fundamental band gap and alignment of two-dimensional semiconductors explored by machine learning[J]. Chinese Physics B, 2020, 29(4): 046101. doi: 10.1088/1674-1056/ab75d5
[39]	TAWFIK S A, ISAYEV O, STAMPFL C, et al. Efficient prediction of structural and electronic properties of hybrid 2D materials using complementary DFT and machine learning approaches[J]. Advanced Theory and Simulations, 2019, 2(1): 1800128. doi: 10.1002/adts.v2.1
[40]	YE W, CHEN C, WANG Z, et al. Deep neural networks for accurate predictions of crystal stability[J]. Nature Communications, 2018, 9(1): 3800. doi: 10.1038/s41467-018-06322-x pmid: 30228262
[41]	WANG H, ZHANG L, HAN J, et al. DeePMD-kit: A deep learning package for many-body potential energy representation and molecular dynamics[J]. Computer Physics Communications, 2018, 228: 178-184. doi: 10.1016/j.cpc.2018.03.016
[42]	PIMACHEV A K, Neogi S. First-principles prediction of electronic transport in fabricated semiconductor heter-ostructures via physics-aware machine learning[J]. npj Computational Materials, 2021, 7(1): 93. doi: 10.1038/s41524-021-00562-0
[43]	PIMACHEV A K, Neogi S. First-principles prediction of electronic transport in fabricated semiconductor heter-ostructures via physics-aware machine learning[J]. npj Computational Materials, 2021, 7(1): 93. doi: 10.1038/s41524-021-00562-0
[44]	XIE Y, STEARRETT R. Machine Learning based CVD Virtual Metrology in Mass Produced Semiconductor Pro-cess[J]. rXiv preprint arXiv:2107.05071, 2021. https://doi.org/10.48550/arXiv.2107.05071. doi: https://doi.org/10.48550/arXiv.2107.05071
[45]	ANTONO E, MATSUZAWA N N, LING J, et al. Mac-hine-learning guided quantum chemical and mole-cular dynamics calculations to design novel hole-condu-cting organic materials[J]. The Journal of Physical Chem-istry A, 2020, 124(40): 8330-8340.
[46]	OWOLABI T O, ABD RAHMAN M A. Prediction of band gap energy of doped graphitic carbon nitride using genetic algorithm-based support vector regression and extreme learning machine[J]. Symmetry, 2021, 13(3): 411. doi: 10.3390/sym13030411
[47]	EL MOURABIT Y, EL HABOUZ Y, ZOUGAGH H, et al. Predictive system of semiconductor failures based on machine learning approach[J]. International journal of advanced computer science and applications (IJACSA), 2020, 11(12): 199-203.
[48]	LIU Z, JIANG M, LUO T. Leverage electron properties to predict phonon properties via transfer learning for semi-conductors[J]. Science advances, 2020, 6(45): eabd1356.
[49]	TSUNOOKA Y, KOKUBO N, HATASA G, et al. High-speed prediction of computational fluid dynamics sim-ulation in crystal growth[J]. CrystEngComm, 2018, 20(41): 6546-6550. doi: 10.1039/C8CE00977E
[50]	YU M, MOAYEDPOUR S, YANG S, et al. Dependence of the electronic structure of the EuS/InAs interface on the bonding configuration[J]. Physical Review Materials, 2021, 5(6): 064606. doi: 10.1103/PhysRevMaterials.5.064606
[51]	HIMANEN L, JÄGER M O J, MOROOKA E V, et al. DS-cribe: Library of descriptors for machine learning in mat-erials science[J]. Computer Physics Communications, 2020, 247: 106949. doi: 10.1016/j.cpc.2019.106949
[52]	CHMIELA S, SAUCEDA H E, POLTAVSKY I, et al. sGDML: Constructing accurate and data efficient molecular force fields using machine learning[J]. Com-puter Physics Communications, 2019, 240: 38-45.

有监督学习算法在材料科学中的应用

Application of Supervised Learning Algorithms in Materials Science

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 4

参考文献 52

相关文章 15

编辑推荐

Metrics

本文评价

[1]	孟哲, 余粟. 基于Spark和优化BP神经网络的出租车需求预测模型[J]. 数据与计算发展前沿, 2023, 5(4): 112-126.
[2]	田一擎, 程曦, 冯博靖. 企业信用评级计算模型综述[J]. 数据与计算发展前沿, 2023, 5(4): 139-153.
[3]	陈美霖, 刘端阳, 徐黎明, 汪洋. 基于机器学习的力场模型研究综述[J]. 数据与计算发展前沿, 2023, 5(4): 27-37.
[4]	陈栋, 李明, 陈淑文. 结合Transformer和多层特征聚合的高光谱图像分类算法[J]. 数据与计算发展前沿, 2023, 5(3): 138-151.
[5]	李妍,何洪波,王闰强. 微博热度预测研究综述[J]. 数据与计算发展前沿, 2023, 5(2): 119-135.
[6]	高添,朱教君,张金鑫,孙一荣,于丰源,滕德雄,卢德亮,于立忠,王宗国. 基于新一代信息技术的温带森林生态系统碳通量精准计量[J]. 数据与计算发展前沿, 2023, 5(2): 60-72.
[7]	王凡,冯立强,曹荣强. 大数据驱动的海洋人工智能服务平台设计与应用[J]. 数据与计算发展前沿, 2023, 5(2): 73-85.
[8]	范少萍,张志强. 数据与技术双轮驱动的生物医学信息学发展与展望[J]. 数据与计算发展前沿, 2023, 5(1): 41-54.
[9]	许淞源,刘峰. ESDRec：一种面向地球大数据平台的数据推荐模型[J]. 数据与计算发展前沿, 2023, 5(1): 55-64.
[10]	杨韬,戴健. 海鸥算法的改进及其在工程设计优化问题中的应用[J]. 数据与计算发展前沿, 2022, 4(6): 129-144.
[11]	童昭,王露笛,朱小杰,杜一. 基于预训练模型的军事领域命名实体识别研究[J]. 数据与计算发展前沿, 2022, 4(5): 120-128.
[12]	赵忠斌,蔡满春,芦天亮. 融合多头注意力机制的网络恶意流量检测[J]. 数据与计算发展前沿, 2022, 4(5): 60-67.
[13]	石雪梅,朱克亮,张祥民,张树涛,陈良锋. 基于生成对抗网络的有遮挡人脸修复方法[J]. 数据与计算发展前沿, 2022, 4(4): 123-131.
[14]	危婷,张宏海,蔺小丽,张蕾蕾,王妍,贾金峰. 云服务网站用户复访行为预测模型研究[J]. 数据与计算发展前沿, 2022, 4(3): 124-130.
[15]	孙永谦,张茹茹,林子涵,张圣林,谭智元,张玉志. KPI异常检测方法评估[J]. 数据与计算发展前沿, 2022, 4(3): 46-65.