Machine learning with systematic density-functional theory calculations: Application to melting temperatures of single- and binary-component solids

被引:245
作者
Seko, Atsuto [1 ]
Maekawa, Tomoya [1 ]
Tsuda, Koji [2 ,3 ]
Tanaka, Isao [1 ,4 ,5 ]
机构
[1] Kyoto Univ, Dept Mat Sci & Engn, Kyoto 6068501, Japan
[2] Japan Sci & Technol Agcy, Minato Discrete Struct Manipulat Syst Project, ERATO, Sapporo, Hokkaido 0600814, Japan
[3] Natl Inst Adv Ind Sci & Technol, Computat Biol Res Ctr, Tokyo 1350064, Japan
[4] Kyoto Univ, Ctr Elements Strategy Initiat Struct Mat ESISM, Kyoto 6068501, Japan
[5] Japan Fine Ceram Ctr, Nanostruct Res Lab, Nagoya, Aichi 4568587, Japan
基金
日本学术振兴会;
关键词
REGRESSION; DESIGN;
D O I
10.1103/PhysRevB.89.054303
中图分类号
T [工业技术];
学科分类号
08 ;
摘要
A combination of systematic density-functional theory (DFT) calculations and machine learning techniques has a wide range of potential applications. This study presents an application of the combination of systematic DFT calculations and regression techniques to the prediction of the melting temperature for single and binary compounds. Here we adopt the ordinary least-squares regression, partial least-squares regression, support vector regression, and Gaussian process regression. Among the four kinds of regression techniques, SVR provides the best prediction. The inclusion of physical properties computed by the DFT calculation to a set of predictor variables makes the prediction better. In addition, limitation of the predictive power is shown when extrapolation from the training dataset is required. Finally, a simulation to find the highest melting temperature toward the efficient materials design using kriging is demonstrated. The kriging design finds the compound with the highest melting temperature much faster than random designs. This result may stimulate the application of kriging to efficient materials design for a broad range of applications.
引用
收藏
页数:9
相关论文
共 37 条
[1]  
Akaike H., 1998, Selected papers of Hirotugu Akaike, P199, DOI DOI 10.1007/978-1-4612-1694-0_15
[2]  
Ambacher O., 1998, J PHYS D, V31, P2653
[3]  
Bishop C., 2006, PATTERN RECOGN, DOI DOI 10.1117/1.2819119
[4]   IMPROVED TETRAHEDRON METHOD FOR BRILLOUIN-ZONE INTEGRATIONS [J].
BLOCHL, PE ;
JEPSEN, O ;
ANDERSEN, OK .
PHYSICAL REVIEW B, 1994, 49 (23) :16223-16233
[5]   Computational screening of perovskite metal oxides for optimal solar light capture [J].
Castelli, Ivano E. ;
Olsen, Thomas ;
Datta, Soumendu ;
Landis, David D. ;
Dahl, Soren ;
Thygesen, Kristian S. ;
Jacobsen, Karsten W. .
ENERGY & ENVIRONMENTAL SCIENCE, 2012, 5 (02) :5814-5819
[6]   Opportunities and challenges for first-principles materials design and applications to Li battery materials [J].
Ceder, Gerbrand .
MRS BULLETIN, 2010, 35 (09) :693-701
[7]   LIBSVM: A Library for Support Vector Machines [J].
Chang, Chih-Chung ;
Lin, Chih-Jen .
ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY, 2011, 2 (03)
[8]   MELTING-POINT TRENDS IN INTERMETALLIC ALLOYS [J].
CHELIKOWSKY, JR ;
ANDERSON, KE .
JOURNAL OF PHYSICS AND CHEMISTRY OF SOLIDS, 1987, 48 (02) :197-205
[9]  
Cherkassky V, 1997, IEEE Trans Neural Netw, V8, P1564, DOI 10.1109/TNN.1997.641482
[10]   Predicting crystal structures with data mining of quantum calculations [J].
Curtarolo, S ;
Morgan, D ;
Persson, K ;
Rodgers, J ;
Ceder, G .
PHYSICAL REVIEW LETTERS, 2003, 91 (13)