Integration and Optimization of Material Data Mining and Machine Learning Tools

doi:10.11871/jfdc.issn.2096-742X.2020.04.009

Abstract

Abstract:

[Objective] Aiming at handling the current situation that there are high barriers impeding materials science researchers to take advantages of machine learning algorithms, this article focuses on developing a user-friendly and highly automated machine learning system for material data mining named Auto-Mat. [Methods] We have integrated some existing methods and machine learning algorithms in MatMiner and scikit-learn, and defined a data dictionary to read data from different material calculation databases. At the same time, we have developed some algorithms for feature selection and processing. [Results] It can provide the system with a visual interaction and display interface for data mining and machine learning modules under a unified data format. With the optimized algorithms, the performances of models are improved. [Limitations] For data acquisition, currently only the data in the MatMiner API can be obtained, and the writing of related code is also fully synchronized with the MatMiner API. So the scalability is poor. Moreover, at present, the execution speed of some core algorithms needs to be improved. [Conclusions] Through this system, users can read data from several mainstream databases such as Materials Project in one shot and quickly build their own material data mining workflow. In the end, a comparative analysis of two cases shows that our platform has a positive effect on reducing the barriers for users to use machine learning methods on material data mining.

Key words: materials science, data mining, visual interactive interface, data summary, feature extraction, simulated annealing algorithm, MatCloud

Dong Jiayuan,Yang Xiaoyu. Integration and Optimization of Material Data Mining and Machine Learning Tools[J]. Frontiers of Data and Computing, 2020, 2(4): 105-120.

Figures/Tables 21

Fig.1

Fig.2

Fig.3

Table 1

Fig.4

Table 2

Fig.5

Table 3

Fig.6

Fig.7

Table 4

Fig.8

Table 5

Fig.9

Fig.10

Fig.11

Table 6

Fig.12

Fig.13

Table 7

Fig.14

References 29

[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]	BLAISZIK B., CHARD K., PRUYNE J ., 等. The Materials Data Facility: Data Services to Advance Materials Science Research[J]. JOM:the Journal of the Minerals Metals & Materials Society, 2016,68(8):2045-2052.
[3]	SWAMI A., JAIN R . Scikit-learn: Machine Learning in Python[J]. Journal of Machine Learning Research, 2012,12(10):2825-2830.
[4]	BLOKHIN E., VILLARS P ., The PAULING FILE Project and Materials Platform for Data Science: From Big Data Toward Materials Genome[M]. Handbook of Materials Modeling. 2018.
[5]	MCKINNEY W . Python for data analysis[M]. 东南大学出版社, 2013.
[6]	O’Mara J, MEREDIG B, MICHEL K . Materials Data Infrastructure: A Case Study of the Citrination Platform to Examine Data Import, Storage, and Access[J]. JOM, 2016,68(8):2031-2034.
[7]	LARSEN A H, MORTENSEN J J, BLOMQVIST J , et al. The atomic simulation environment—a Python library for working with atoms[J]. Journal of Physics: Condensed Matter, 2017,29(27):273002. doi: 10.1088/1361-648X/aa680e pmid: 28323250
[8]	MONTAVON G., HANSEN K., FAZLI S ., et al. Learning Invariant Representations of Molecules for Atomization Energy Prediction [C]. International Conference on Neural Information Processing Systems. Curran Associates Inc. 2012.
[9]	RUPP M . Many-Body Tensor Representation for Machine Learning of Materials [C]. Aps March Meeting. APS March Meeting Abstracts, 2017.
[10]	WARD L., DUNN A., FAGHANINIA A ., 等. Matminer: An open source toolkit for materials data mining[J]. Computational Materials Science, 2018,152:60-69.
[11]	SHEN C., BAO X., TAN J ., 等. Two noise-robust axial scanning multi-image phase retrieval algorithms based on Pauta criterion and smoothness constraint[J]. Optics Express, 2017,25(14):16235. doi: 10.1364/OE.25.016235 pmid: 28789131
[12]	WILHELM J, FREY E . Radial Distribution Function of Semiflexible Polymers[J]. Physical Review Letters, 1996,77(12):2581. doi: 10.1103/PhysRevLett.77.2581 pmid: 10061990
[13]	STEINHARDT P J, NELSON D R, Ronchetti M . Bond-Orientational Order in Liquids and Glasses[J]. Physical review. B, Condensed matter, 1983,28(2):784-805.
[14]	RATOWSKY R P, FLECK J A . Treatment of angular derivatives in the Schrödinger equation by the finite Fourier series method[J]. Journal of Computational Physics, 1991,89(2):490-490.
[15]	GOH G B, HODAS N O, VISHNU A . Deep learning for computational chemistry[J]. Journal of Computational Chemistry, 2017,38(16):1291-1307. doi: 10.1002/jcc.24764 pmid: 28272810
[16]	RUPP M, TKATCHENKO A MÜLLER, KLAUS-ROBERT , et al. Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning[J]. PHYSICAL REVIEW LETTERS, 2012,108(5):58301-0.
[17]	LING J, HUTCHINSON M, ANTONO E , et al. High-Dimensional Materials and Process Optimization Using Data-Driven Experimental Design with Well-Calibrated Uncertainty Estimates[J]. Integrating Materials and Manufacturing Innovation, 2017,6:207-217.
[18]	KOHAVI R . A Study of Cross-Validation and Bootstrap for Accuracy Estimation and Model Selection [C]. International joint conference on Artificial intelligence. Morgan Kaufmann Publishers Inc. 1995.
[19]	OLSTHOORN B, GEILHUFE R M, BORYSOV S S , et al. Band Gap Prediction for Large Organic Crystal Structures with Machine Learning[J]. Advanced Quantum Technologies, 2019,2:7-8.
[20]	ONG S P, RICHARDS W D, JAIN A , et al. Python Materials Genomics (pymatgen): A robust, open-source python library for materials analysis[J]. Computational Materials Science, 2013,68:314-319. doi: 10.1016/j.commatsci.2012.10.028
[21]	KAY H. F., BAILEY P. C ., Structure and Properties of CaTiO3[J]. Acta Crystallographica, 1957,10(3):219-226. doi: 10.1107/S0365110X57000675
[22]	OUYANG R, CURTAROLO S, AHMETCIK E , et al. SISSO: A compressed-sensing method for identifying the best low-dimensional descriptor in an immensity of offered candidates[J]. Physical Review Materials, 2018,2(8):083802. doi: 10.1103/PhysRevMaterials.2.083802
[23]	MITRA P., MURTHY C.A., PAL S. K . Unsupervised feature selection using feature similarity[J]. Pattern Analysis & Machine Intelligence IEEE Transactions on, 2002,24(3):301-312.
[24]	HARTIGAN J.A., WONG M.A . A K-means clustering algorithm[J]. Appl Stat, 2013,28(1):100-108. doi: 10.2307/2346830
[25]	Manuel Arellano and Stephen Bond. Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations[J]. Review of Economic Studies, 58(2):277-297. doi: 10.2307/2297968
[26]	KIRKLIN, SCOTT, SAAL, JAMES E, MEREDIG, BRYCE , 等. The Open Quantum Materials Database (OQMD): assessing the accuracy of DFT formation energies[J]. npj Computational Materials, 1:15010. doi: 10.1038/npjcompumats.2015.10
[27]	SILVER D., HUANG A., MADDISON C. J., GUEZ A., SIFRE L., VAN DEN DRIESSCHE, G., … & DIELEMAN S . Mastering the game of Go with deep neural networks and tree search[J]. Nature, 2016,529(7587):484-489. doi: 10.1038/nature16961 pmid: 26819042
[28]	YANG X, WANG Z , et al. MatCloud: A high-throughput computational infrastructure for integrated management of materials simulation, data and resources[J]. Computational Materials Science, 2018,146:319-333. doi: 10.1016/j.commatsci.2018.01.039
[29]	YANG X, WANG Z . et al. MatCloud, a high-throughput computational materials infrastructure: Present, future visions, and challenges[J]. 中国物理:英文版, 027(011):104-111.

评价指标	对照组（不使用重组算法）	特征重组
RMSE	1.38	0.78
R2-Score	0.41	0.81
Pearson Coefficients	0.77	0.90
Spearman Coefficients	0.54	0.77

评价指标	对照组（默认参数）	超参数优化
RMSE	1.38	0.82
R2-Score	0.41	0.67
Pearson Coefficients	0.77	0.85
Spearman Coefficients	0.54	0.70

特征名称	与目标值的相关系数
mean wt CN_2	-0.1256
max packing efficiency.1	-0.1197
wt CN_6	-0.1859
mean row	-0.0046
compound possible	0.4690
std_dev row	0.0136
bond #4	-0.0841
fraction electrons	-0.1233
std_dev wt CN_1	0.3843
maximum GSmagmom	0.0050

特征名称	与目标值的相关系数
Descriptor-1	0.79
Descriptor-2	0.82

特征名称	与目标值的相关系数
range NUnfilled	0.4514
minimum local difference	-0.1821
mean Column	-0.0546
bond #0	0.0412
bond #3	0.2903
bond #4	0.3283
local difference in MendeleevNumber	-0.0359
std_dev melting_point	0.1888
mean local difference in MeltingT	0.2480
bond #2	0.2513