Hostname: page-component-77f85d65b8-g4pgd Total loading time: 0 Render date: 2026-04-18T15:24:35.253Z Has data issue: false hasContentIssue false
  • English
  • Français

Embedding domain knowledge for machine learning of complex material systems

Published online by Cambridge University Press:  10 July 2019

Christopher M. Childs  and
Newell R. Washburn
Christopher M. Childs
Affiliation:
Washburn Laboratory, Department of Chemistry, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213, USA
Newell R. Washburn*
Affiliation:
Department of Chemistry, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213, USA Department of Biomedical Engineering, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213, USA
*
Address all correspondence to Newell R. Washburn at washburn@andrew.cmu.edu
Get access

Abstract

Machine learning (ML) has revolutionized disciplines within materials science that have been able to generate sufficiently large datasets to utilize algorithms based on statistical inference, but for many important classes of materials the datasets remain small. However, a rapidly growing number of approaches to embedding domain knowledge of materials systems are reducing data requirements and allowing broader applications of ML. Furthermore, these hybrid approaches improve the interpretability of the predictions, allowing for greater physical insights into the factors that determine material properties. This review introduces a number of these strategies, providing examples of how they were implemented in ML algorithms and discussing the materials systems to which they were applied.

Information

Type
Artificial Intelligence Prospectives
Information
MRS Communications , Volume 9 , Issue 3 , September 2019 , pp. 806 - 820
Copyright
Copyright © The Author(s) 2019 

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Article purchase

Temporarily unavailable

References

1.Kittel, C.: Physical theory of ferromagnetic domains. Rev. Mod. Phys. 21, 541 (1949).Google Scholar
2.Flory, P.J.: Molecular theory of rubber elasticity. Polym. J. 17, 1 (1985).Google Scholar
3.Stickel, J.J. and Powell, R.L.: Fluid mechanics and rheology of dense suspensions. Annu. Rev. Fluid Mech. 37, 129 (2005).Google Scholar
4.DeCost, B.L., Francis, T., and Holm, E.A.: Exploring the microstructure manifold: image texture representations applied to ultrahigh carbon steel microstructures. Acta Mater 133, 30 (2017).Google Scholar
5.Saravanan, K., Kitchin, J.R., von Lilienfeld, O.A., and Keith, J.A.: Alchemical predictions for computational catalysis: potential and limitations. J. Phys. Chem. Lett. 8, 5002 (2017).Google Scholar
6.Ramprasad, R., Batra, R., Pilania, G., Mannodi-Kanakkithodi, A., and Kim, C.: Machine learning in materials informatics: recent applications and prospects. NPJ Comput. Mater. 3, 54 (2017).Google Scholar
7.Jain, A., Ong, S.P., Hautier, G., Chen, W., Richards, W.D., Dacek, S., Cholia, S., Gunter, D., Skinner, D., Ceder, G., and Persson, K.A.: Commentary: The Materials Project: a materials genome approach to accelerating materials innovation. APL Mater. 1, 011002 (2013).Google Scholar
8.McDowell, D.L. and Kalidindi, S.R.: The materials innovation ecosystem: a key enabler for the Materials Genome Initiative. MRS Bull. 41, 326 (2016).Google Scholar
9.Qin, M., Lin, Z., Wei, Z., Zhu, B., Yuan, J., Takeuchi, I., and Jin, K.: High-throughput research on superconductivity. Chinese Phys. B 27, 127402 (2018).Google Scholar
10.Gani, T.Z.H. and Kulik, H.J.: Understanding and breaking scaling relations in single-site catalysis: Methane to methanol conversion by Fe IV O. ACS Catal. 8, 975 (2018).Google Scholar
11.Ramakrishna, S., Zhang, T.Y., Lu, W.-C., Qian, Q., Low, J.S.C., Yune, J.H.R., Tan, D.Z.L., Bressan, S., Sanvito, S., and Kalidindi, S.R.: Materials informatics. J. Intell. Manuf (2018). https://doi.org/10.1007/s10845-018-1392-0Google Scholar
12.McBride, M., Persson, N., Reichmanis, E., Grover, M., McBride, M., Persson, N., Reichmanis, E., and Grover, M.A.: Solving materials’ small data problem with dynamic experimental databases. Processes 6, 79 (2018).Google Scholar
13.Kuhne, R., Ebert, R.-U., and Schuurmann, G.: Model selection based on structural similarity-method description and application to water solubility prediction. J. Chem. Inf. Model. 46, 636 (2006).Google Scholar
14.Hughes, L.D., Palmer, D.S., Nigsch, F., and Mitchell, J.B.O.: Why are some properties more difficult to predict than others? A study of QSPR models of solubility, melting point, and log P. J. Chem. Inf. Model. 48, 220 (2008).Google Scholar
15.Sanchez-Lengeling, B., Roch, L.M., Perea, J.D., Langner, S., Brabec, C.J., and Aspuru-Guzik, A.: A Bayesian approach to predict solubility parameters. Adv. Theory Simul 2, 1 (2019).Google Scholar
16.Meredig, B., Agrawal, A., Kirklin, S., Saal, J.E., Doak, J.W., Thompson, A., Zhang, K., Choudhary, A., and Wolverton, C.: Combinatorial screening for new materials in unconstrained composition space with machine learning. Phys. Rev. B 89, 094104 (2014).Google Scholar
17.Hansen, K., Biegler, F., Ramakrishnan, R., Pronobis, W., von Lilienfeld, O.A., Müller, K.-R., and Tkatchenko, A.: Machine learning predictions of molecular properties: accurate many-body potentials and nonlocality in chemical space. J. Phys. Chem. Lett. 6, 2326 (2015).Google Scholar
18.Liu, Y., Zhao, T., Ju, W., and Shi, S.: Materials discovery and design using machine learning. J. Mater. 3, 159 (2017).Google Scholar
19.Rowe, R.C. and Colbourn, E.A.: Neural computing in product formulation. Chem. Educ. 8, 1 (2003).Google Scholar
20.Tanco, M., Viles, E., Ilzarbe, L., and Alvarez, M.J.: Implementation of design of experiments projects in industry. Appl. Stoch. Model. Bus. Ind. 25, 478 (2009).Google Scholar
21.Montgomery, D.C.: Design and Analysis of Experiments. 8th ed. (Wiley, New York, 2012).Google Scholar
22.Jordan, M.I. and Mitchell, T.M.: Machine learning: trends, perspectives, and prospects. Science 349, 255 (2015).Google Scholar
23.Haenssle, H.A., Fink, C., Schneiderbauer, R., Toberer, F., Buhl, T., Blum, A., Kalloo, A., Ben Hadj Hassen, A., Thomas, L., Enk, A., Uhlmann, L., and Holger Haenssle, m.A.: Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists. Ann. Oncol. 29, 1836 (2018).Google Scholar
24.Griffiths, T.L., Baraff, E.R., and Tenenbaum, J.B.: Using physical theories to infer hidden causal structure. Proc. Annu. Meet. Cogn. Sci. Soc. 26, 500 (2004).Google Scholar
25.Michalski, R.S.: Toward a Unified Theory of Learning: An Outline of Basic Ideas. In First World Conference on the Fundamentals of Artificial Intelligence (Paris, 1991).Google Scholar
26.Carbonell, J.G., Michalski, R.S., and Mitchell, T.M.: An overview of machine learning. In Machine Learning: An Artificial Intelligence Approach, edited by Michalski, R.S., Carbonell, J.G. and Mitchell, T.M. (Springer-Verlag, Berlin, 1983).Google Scholar
27.Tenenbaum, J.B., Griffiths, T.L., and Kemp, C.: Theory-based Bayesian models of inductive learning and reasoning. Trends Cogn. Sci. 10, 309 (2006).Google Scholar
28.Lake, B.M., Salakhutdinov, R., and Tenenbaum, J.B.: Human-level concept learning through probabilistic program induction. Science 350, 1332 (2015).Google Scholar
29.Frawley, W.J. and Piatetsky-Shapior, G.: Knowedge Discovery in Databases. 1st ed. (The MIT Press, Cambridge, 1991).Google Scholar
30.Sacha, D., Sedlmair, M., Zhang, L., Lee, J.A., Peltonen, J., Weiskopf, D., North, S.C., and Keim, D.A.: What you see is what you can change: human-centered machine learning by interactive visualization. Neurocomputing 268, 164 (2017).Google Scholar
31.Jain, A., Hautier, G., Ping Ong, S., and Persson, K.: New opportunities for materials informatics: resources and data mining techniques for uncovering hidden relationships. J. Mater. Res. 31, 977 (2016).Google Scholar
32.Wu, Q., Suetens, P., and Oosterlinck, A.: Integration of heuristic and Bayesian approaches in a pattern-classification system. In Knowledge Discovery Databases, 1st ed, edited by Piatetsky-Shapiro, G., and Frawley, W.J. (The MIT Press, Cambridge, 1991), pp. 249–260.Google Scholar
33.Tibshirani, R.: Regression shrinkage and selection via the Lasso. J. R. Stat. Soc. Ser. B 58, 267 (1996).Google Scholar
34.Mitchell, J.B.O.: Machine learning methods in chemoinformatics. Wiley Interdiscip. Rev. Comput. Mol. Sci. 4, 468 (2014).Google Scholar
35.Mooney, C.Z. and Duval, R.D.: Bootstrapping A Nonparametric Approach to Statistical Inference (Sage Publications, Inc, Newbury Park, CA, 1993).Google Scholar
36.Svetnik, V., Liaw, A., Tong, C., Culberson, J.C., Sheridan, R.P., and Feuston, B.P.: Random forest: a classification and regression tool for compound classification and QSAR modeling. J. Chem. Inf. Comput. Sci. 43, 1947 (2003).Google Scholar
37.Xu, M., Watanachaturaporn, P., Varshney, P.K., and Arora, M.K.: Decision tree regression for soft classification of remote sensing data. Remote Sens. Environ. 97, 322 (2005).Google Scholar
38.Liaw, A. and Wiener, M.: Classification and regression by RandomForest. R News 2/3, 18 (2002).Google Scholar
39.Rasmussen, C.E.: Gaussian processes in machine learning. In Adv. Lect. Mach. Learn. edited by Bousquet, O., von Luxburg, U. and Rätsch, G. (Springer-Verlag, Berlin, 2003), pp. 63–71.Google Scholar
40.Rasmussen, C.E. and Williams, C.K.I.: Gaussian Processes for Machine Learning, 2nd ed. (MIT Press, Cambridge, 2006).Google Scholar
41.Li, H., Collins, C., Tanha, M., Gordon, G.J., and Yaron, D.J.: A density functional tight binding layer for deep learning of chemical hamiltonians. J. Chem. Theory Comput. 14, 5764 (2018).Google Scholar
42.Li, Y., Li, H., Pickard, F.C., Narayanan, B., Sen, F.G., Chan, M.K.Y., Sankaranarayanan, S.K.R.S., Brooks, B.R., and Roux, B.: Machine learning force field parameters from ab initio data. J. Chem. Theory Comput 13, 4492 (2017).Google Scholar
43.Schütt, K.T., Glawe, H., Brockherde, F., Sanna, A., Müller, K.R., and Gross, E.K.U.: How to represent crystal structures for machine learning: towards fast prediction of electronic properties. Phys. Rev. B 89, 205118 (2014).Google Scholar
44.Hu, L., Wang, X., Wong, L., and Chen, G.: Combined first-principles calculation and neural-network correction approach for heat of formation. J. Chem. Phys. 119, 11501 (2003).Google Scholar
45.von Lilienfeld, O.A.: Quantum machine learning in chemical compound space. Angew. Chemie Int. Ed. 57, 4164 (2018).Google Scholar
46.Gardas, R.L. and Coutinho, J.A.P.: A group contribution method for viscosity estimation of ionic liquids. Fluid Phase Equilib. 266, 195 (2008).Google Scholar
47.Paduszyński, K. and Domańska, U.: Viscosity of ionic liquids: an extensive database and a new group contribution model based on a feed-forward artificial neural network. J. Chem. Inf. Model. 54, 1311 (2014).Google Scholar
48.Mehrkesh, A. and Karunanithi, A.T.: New quantum chemistry-based descriptors for better prediction of melting point and viscosity of ionic liquids. Fluid Phase Equilib. 427, 498 (2016).Google Scholar
49.Preiss, U., Bulut, S., and Krossing, I.: In silico prediction of the melting points of ionic liquids from thermodynamic considerations. A case study on 67 salts with a melting point range of 337 °C. J. Phys. Chem. B 114, 11133 (2010).Google Scholar
50.Fatehi, M.-R., Raeissi, S., and Mowla, D.: Estimation of viscosities of pure ionic liquids using an artificial neural network based on only structural characteristics. J. Mol. Liq. 227, 309 (2017).Google Scholar
51.Kalidindi, S.R. and De Graef, M.: Materials data science: current status and future outlook. Annu. Rev. Mater. Res. 45, 171 (2015).Google Scholar
52.Magnan, C.N. and Baldi, P.: SSpro/ACCpro 5: almost perfect prediction of protein secondary structure and relative solvent accessibility using profiles, machine learning and structural similarity. Bioinformatics 30, 2592 (2014).Google Scholar
53.Pilania, G., Wang, C., Jiang, X., Rajasekaran, S., and Ramprasad, R.: Accelerating materials property predictions using machine learning. Sci. Rep. 3, 2810 (2013).Google Scholar
54.Vandenburg, H.J., Clifford, A.A., Bartle, K.D., Carlson, R.E., Carroll, J., and Newton, I.D.: A simple solvent selection method for accelerated solvent extraction of additives from polymers. Analyst 124, 1707 (1999).Google Scholar
55.Hansen, C.: Hansen Solubility Parameters - A User's Handbook (CRC Press, Boca Raton, 1999).Google Scholar
56.Lindvig, T., Michelsen, M.L., and Kontogeorgis, G.M.: A Flory – Huggins model based on the Hansen solubility parameters. Fluid Phase Equilib. 203, 247 (2002).Google Scholar
57.Albahri, T.A.: Accurate prediction of the solubility parameter of pure compounds from their molecular structures. Fluid Phase Equilib. 379, 96 (2014).Google Scholar
58.Stefanis, E. and Panayiotou, C.: Prediction of Hansen solubility parameters with a new group-contribution method. Int. J. Thermophys. 29, 568 (2008).Google Scholar
59.Gal, Y. and Ghahramani, Z.: Proceeding of 33rd International Conference on Machine Learning (New York, 2016).Google Scholar
60.Cao, L., Li, C., and Mueller, T.: The use of cluster expansions to predict the structures and properties of surfaces and nanostructured materials. J. Chem. Inf. Model. 58, 2401 (2018).Google Scholar
61.Mueller, T. and Ceder, G.: Bayesian approach to cluster expansions. Phys. Rev. B 80, 024103 (2009).Google Scholar
62.Butler, K.T., Davies, D.W., Cartwright, H., Isayev, O., and Walsh, A.: Machine learning for molecular and materials science. Nature 559, 547 (2018).Google Scholar
63.Ling, J., Jones, R., and Templeton, J.: Machine learning strategies for systems with invariance properties. J. Comput. Phys. 318, 22 (2016).Google Scholar
64.E, W. and Ming, P.: Cauchy–Born rule and the stability of crystalline solids: static problems. Arch. Ration. Mech. Anal 183, 241 (2007).Google Scholar
65.Cireşan, D.C., Meier, U., Gambardella, L.M., and Schmidhuber, J.: Deep, big, simple neural nets for handwritten digit recognition. Neural Comput. 22, 3207 (2010).Google Scholar
66.Kambouchev, N., Fernandez, J., and Radovitzky, R.: A polyconvex model for materials with cubic symmetry. Model. Simul. Mater. Sci. Eng. 15, 451 (2007).Google Scholar
67.Karpatne, A., Atluri, G., Faghmous, J.H., Steinbach, M., Banerjee, A., Ganguly, A., Shekhar, S., Samatova, N., and Kumar, V.: Theory-guided data science: a new paradigm for scientific discovery from data. IEEE Trans. Knowl. Data Eng. 29, 2318 (2017).Google Scholar
68.Xiao, H., Wu, J.-L., Wang, J.-X., Sun, R., and Roy, C.J.: Quantifying and reducing model-form uncertainties in Reynolds-averaged Navier–Stokes simulations: a data-driven, physics-informed Bayesian approach. J. Comput. Phys 324, 115 (2016).Google Scholar
69.Wang, J.-X., Wu, J.-L., and Xiao, H.: Physics-informed machine learning approach for reconstructing Reynolds stress modeling discrepancies based on DNS data. Phys. Rev. Fluids 2, 34603 (2017).Google Scholar
70.Ghiringhelli, L.M., Vybiral, J., Levchenko, S.V., Draxl, C., and Scheffler, M.: Big data of materials science: critical role of the descriptor. Phys. Rev. Lett. 114, 105503 (2015).Google Scholar
71.Menon, A., Gupta, C., Perkins, K.M., DeCost, B.L., Budwal, N., Rios, R.T., Zhang, K., Póczos, B., and Washburn, N.R.: Elucidating multi-physics interactions in suspensions for the design of polymeric dispersants: a hierarchical machine learning approach. Mol. Syst. Des. Eng. 2, 263 (2017).Google Scholar
72.Hirata, T., Ye, J., Branicio, P., Zheng, J., Lange, A., Plank, J., and Sullivan, M.: Adsorbed conformations of PCE superplasticizers in cement pore solution unraveled by molecular dynamics simulations. Sci. Rep. 7, 16599 (2017).Google Scholar
73.Marchon, D., Juilland, P., Gallucci, E., Frunz, L., and Flatt, R.J.: Molecular and submolecular scale effects of comb-copolymers on tri-calcium silicate reactivity: toward molecular design. J. Am. Ceram. Soc. 100, 817 (2016).Google Scholar
74.Ding, J.-T. and Li, Z.: Effects of Metakaolin and silica fume on properties of concrete. ACI Mater. J. 99, 393 (2002).Google Scholar
75.Washburn, N.R., Menon, A., Childs, C.M., Poczos, B., and Kurtis, K.E.: Machine learning approaches to admixture design for clay-based cements. In Calcined Clays for Sustainable Concrete, edited by Martirena, F., Favier, A. and Scrivener, K. (Springer, Dordrecht, 2017), pp. 488–493.Google Scholar
76.Menon, A., Childs, C.M., Poczós, B., Washburn, N.R., and Kurtis, K.E.: Molecular engineering of superplasticizers for Metakaolin-Portland cement blends with hierarchical machine learning. Adv. Theory Simul 2, 1800164 (2018).Google Scholar
77.Yoshioka, K., Sakai, E., Daimon, M., and Kitahara, A.: Role of steric hindrance in the performance of superplasticizers for concrete. J. Am. Ceram. Soc. 80, 2667 (1997).Google Scholar
78.Hutchinson, M.L., Antono, E., Gibbons, B.M., Paradiso, S., Ling, J., and Meredig, B.: Overcoming data scarcity with transfer learning. In 31st Conference on Neural Information Processing Systems (NIPS 2017) (Long Beach, 2017), pp. 1–10.Google Scholar
79.Welborn, M., Cheng, L., and Miller, T.F.: Transferability in machine learning for electronic structure via the molecular orbital basis. J. Chem. Theory Comput. 14, 4772 (2018).Google Scholar
80.Bartók, A.P., De, S., Poelking, C., Bernstein, N., Kermode, J.R., Csányi, G., and Ceriotti, M.: Machine learning unifies the modeling of materials and molecules. Sci. Adv. 3, e1701816 (2017).Google Scholar
81.Parish, E.J. and Duraisamy, K.: A paradigm for data-driven predictive modeling using field inversion and machine learning. J. Comput. Phys. 305, 758 (2016).Google Scholar