Exploring model complexity in machine learned potentials for simulated properties
Source abstract
Abstract
Abstract Machine learning (ML) enables the development of interatomic potentials with the accuracy of first principles methods while retaining the speed and parallel efficiency of empirical potentials. While ML potentials traditionally use atom-centered descriptors as inputs, different models such as linear regression and neural networks map descriptors to atomic energies and forces. This begs the question: what is the improvement in accuracy due to model complexity irrespective of descriptors? We curate three datasets to investigate this question in terms of ab initio energy and force errors: (1) solid and liquid silicon, (2) gallium nitride, and (3) the superionic conductor Li$$10$$Ge(PS$$6$$)$$2$$ (LGPS). We further investigate how these errors affect simulated properties and verify if the improvement in fitting errors corresponds to measurable improvement in property prediction. By assessing different models, we observe correlations between fitting quantity (e.g. atomic force) error and simulated property error with respect to ab initio values. Graphical abstract
Source: crossref · Quality: candidate unverified · Source record ↗
Research context
OpenAlex topics: Machine Learning in Materials Science · Fuel Cells and Related Materials · X-ray Diffraction in Crystallography
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (47)
Show source-provided references
1. K. Choudhary, B. DeCost, C. Chen, A. Jain, F. Tavazza, R. Cohn, C.W. Park, A. Choudhary, A. Agrawal, S.J. Billinge et al., Recent advances and applications of deep learning methods in materials science. npj Comput. Mater. 8(1), 59 (2022)10.1038/s41524-022-00734-6 ↗
2. H. Tafrishi, S. Sadeghzadeh, R. Ahmadi, Molecular dynamics simulations of phase change materials for thermal energy storage: a review. RSC Adv. 12(23), 14776–14807 (2022)10.1039/D2RA02183H ↗
3. N. Yao, X. Chen, Z.-H. Fu, Q. Zhang, Applying classical, ab initio, and machine-learning molecular dynamics simulations to the liquid electrolyte for rechargeable batteries. Chem. Rev. 122(12), 10970–11021 (2022)10.1021/acs.chemrev.1c00904 ↗
4. J. Vandermause, Y. Xie, J.S. Lim, C.J. Owen, B. Kozinsky, Active learning of reactive Bayesian force fields applied to heterogeneous catalysis dynamics of h/pt. Nature Commun. 13(1), 5183 (2022)10.1038/s41467-022-32294-0 ↗
5. Q. Bai, S. Liu, Y. Tian, T. Xu, A.J. Banegas-Luna, H. Pérez-Sánchez, J. Huang, H. Liu, X. Yao, Application advances of deep learning methods for de novo drug design and molecular dynamics simulation. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 12(3), 1581 (2022)
6. V.L. Deringer, M.A. Caro, G. Csányi, Machine learning interatomic potentials as emerging tools for materials science. Adv. Mater. 31(46), 1902765 (2019)10.1002/adma.201902765 ↗
7. Y. Mishin, Machine-learning interatomic potentials for materials science. Acta Mater. 214, 116980 (2021)10.1016/j.actamat.2021.116980 ↗
8. A.P. Bartók, R. Kondor, G. Csányi, On representing chemical environments. Phys. Rev. B 87(18), 184115 (2013)10.1103/PhysRevB.87.184115 ↗
9. A.P. Thompson, L.P. Swiler, C.R. Trott, S.M. Foiles, G.J. Tucker, Spectral neighbor analysis method for automated generation of quantum-accurate interatomic potentials. J. Comput. Phys. 285, 316–330 (2015)10.1016/j.jcp.2014.12.018 ↗
10. R. Drautz, Atomic cluster expansion for accurate and transferable interatomic potentials. Phys. Rev. B 99(1), 014104 (2019)10.1103/PhysRevB.99.014104 ↗
11. J. Behler, M. Parrinello, Generalized neural-network representation of high-dimensional potential-energy surfaces. Phys. Rev. Lett. 98(14), 146401 (2007)10.1103/PhysRevLett.98.146401 ↗
12. N.-C. Nguyen, A. Rohskopf, Proper orthogonal descriptors for efficient and accurate interatomic potentials. J. Comput. Phys. 480, 112030 (2023)10.1016/j.jcp.2023.112030 ↗
13. C.W. Park, M. Kornbluth, J. Vandermause, C. Wolverton, B. Kozinsky, J.P. Mailoa, Accurate and scalable graph neural network force field and molecular dynamics with direct force architecture. npj Comput. Mater. 7(1), 73 (2021)10.1038/s41524-021-00543-3 ↗
14. A. Zhu, S. Batzner, A. Musaelian, B. Kozinsky, Fast uncertainty estimates in deep learning interatomic potentials. Preprint at http://arxiv.org/abs/2211.09866 (2022)10.1063/5.0136574 ↗
15. A. Rohskopf, C. Sievers, N. Lubbers, M. Cusentino, J. Goff, J. Janssen, M. McCarthy, D.M.O. Zapiain, S. Nikolov, K. Sargsyan, D. Sema, E. Sikorski, L. Williams, A. Thompson, M. Wood, FitSNAP: atomistic machine learning with LAMMPS. J. Open Source Softw. 8(84), 5118 (2023)10.21105/joss.05118 ↗
16. A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, et al., Pytorch: an imperative style, high-performance deep learning library. Adv. Neural Inf. Process. Syst. 32 (2019)
17. J. Bradbury, R. Frostig, P. Hawkins, M.J. Johnson, C. Leary, D. Maclaurin, G. Necula, A. Paszke, J. VanderPlas, S. Wanderman-Milne, et al., Jax: composable transformations of python+ numpy programs (2018)
18. M.A. Cusentino, M.A. Wood, A.P. Thompson, Explicit multielement extension of the spectral neighbor analysis potential for chemically complex systems. J. Phys. Chem. A 124(26), 5456–5464 (2020)10.1021/acs.jpca.0c02450 ↗
19. M.A. Wood, A.P. Thompson, Extending the accuracy of the snap interatomic potential form. J. Chem. Phys. 148(24), 241721 (2018)10.1063/1.5017641 ↗
20. Y. Zuo, C. Chen, X. Li, Z. Deng, Y. Chen, J. Behler, G. Csányi, A.V. Shapeev, A.P. Thompson, M.A. Wood et al., Performance and cost assessment of machine learning interatomic potentials. J. Phys. Chem. A 124(4), 731–745 (2020)10.1021/acs.jpca.9b08723 ↗
21. K. Hornik, M. Stinchcombe, H. White, Multilayer feedforward networks are universal approximators. Neural Netw. 2(5), 359–366 (1989)10.1016/0893-6080(89)90020-8 ↗
22. B.C. Csáji et al., Approximation with artificial neural networks. Faculty of Sciences, Etvs Lornd University 24(48), 7 (2001)
23. Y. Lysogorskiy, C.V.D. Oord, A. Bochkarev, S. Menon, M. Rinaldi, T. Hammerschmidt, M. Mrovec, A. Thompson, G. Csányi, C. Ortner et al., Performant implementation of the atomic cluster expansion (pace) and application to copper and silicon. npj Comput. Mater. 7(1), 97 (2021)10.1038/s41524-021-00559-9 ↗
24. A.P. Bartók, J. Kermode, N. Bernstein, G. Csányi, Machine learning a general-purpose interatomic potential for silicon. Phys. Rev. X 8(4), 041048 (2018)
25. M.A. Wood, M.A. Cusentino, B.D. Wirth, A.P. Thompson, Data-driven material models for atomistic simulation. Phys. Rev. B 99(18), 184305 (2019)10.1103/PhysRevB.99.184305 ↗
26. S. Nikolov, M.A. Wood, A. Cangi, J.-B. Maillet, M.-C. Marinica, A.P. Thompson, M.P. Desjarlais, J. Tranchida, Data-driven magneto-elastic predictions with scalable classical spin-lattice dynamics. npj Comput. Mater. 7(1), 153 (2021)10.1038/s41524-021-00617-2 ↗
27. K. Nguyen-Cong, J.T. Willman, S.G. Moore, A.B. Belonoshko, R. Gayatri, E. Weinberg, M.A. Wood, A.P. Thompson, I.I. Oleynik, Billion atom molecular dynamics simulations of carbon at extreme conditions and experimental time and length scales. In: Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, pp. 1–12 (2021)10.1145/3458817.3487400 ↗
28. E. Sikorski, M. Cusentino, M. McCarthy, J. Tranchida, M. Wood, A. Thompson, Machine learned interatomic potential for dispersion strengthened plasma facing components. Preprint at http://arxiv.org/abs/2212.01432 (2022)10.1063/5.0135269 ↗
29. D.F. Toit, V.L. Deringer, Cross-platform hyperparameter optimization for machine learning interatomic potentials. J. Chem. Phys. 10(1063/5), 0155618 (2023)
30. D. Oca Zapiain, M.A. Wood, N. Lubbers, C.Z. Pereyra, A.P. Thompson, D. Perez, Training data selection for accuracy and transferability of interatomic potentials. npj Comput. Mater. 8(1), 189 (2022)10.1038/s41524-022-00872-x ↗
31. X. Fu, Z. Wu, W. Wang, T. Xie, S. Keten, R. Gomez-Bombarelli, T. Jaakkola, Forces are not enough: Benchmark and critical evaluation for machine learning force fields with molecular simulations. Preprint at http://arxiv.org/abs/2210.07237 (2022)
32. J. Tsao, S. Chowdhury, M. Hollis, D. Jena, N. Johnson, K. Jones, R. Kaplar, S. Rajan, C. Walle, E. Bellotti et al., Ultrawide-bandgap semiconductors: research opportunities and challenges. Adv. Electron. Mater. 4(1), 1600501 (2018)10.1002/aelm.201600501 ↗
33. E. Minamitani, M. Ogura, S. Watanabe, Simulating lattice thermal conductivity in semiconducting materials using high-dimensional neural network potential. Appl. Phys. Express 12(9), 095001 (2019)10.7567/1882-0786/ab36bc ↗
34. X. Gao, F. Ramezanghorbani, O. Isayev, J.S. Smith, A.E. Roitberg, Torchani: a free and open source pytorch-based deep learning implementation of the ani neural network potentials. J. Chem. Inf. Model. 60(7), 3408–3415 (2020)10.1021/acs.jcim.0c00451 ↗
35. A. Rohskopf, S. Wyant, K. Gordiz, H.R. Seyf, M.G. Muraleedharan, A. Henry, Fast & accurate interatomic potentials for describing thermal vibrations. Comput. Mater. Sci. 184, 109884 (2020)10.1016/j.commatsci.2020.109884 ↗
36. S. Batzner, A. Musaelian, L. Sun, M. Geiger, J.P. Mailoa, M. Kornbluth, N. Molinari, T.E. Smidt, B. Kozinsky, E (3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials. Nature commun. 13(1), 2453 (2022)10.1038/s41467-022-29939-5 ↗
37. J.B. Boyce, B.A. Huberman, Superionic conductors: transitions, structures, dynamics. Phys. Rep. 51(4), 189–265 (1979)10.1016/0370-1573(79)90067-X ↗
38. Y. Kato, S. Hori, R. Kanno, Li10gep2s12-type superionic conductors: synthesis, structure, and ionic transportation. Adv. Energy Mater. 10(42), 2002153 (2020)10.1002/aenm.202002153 ↗
39. X. He, Q. Bai, Y. Liu, A.M. Nolan, C. Ling, Y. Mo, Crystal structural framework of lithium super-ionic conductors. Adv. Energy Mater. 9(43), 1902078 (2019)10.1002/aenm.201902078 ↗
40. G. Winter, R. Gómez-Bombarelli, Simulations with machine learning potentials identify the ion conduction mechanism mediating non-Arrhenius behavior in lgps. Preprint at http://arxiv.org/abs/2211.05713 (2022)10.1088/2515-7655/acbbef ↗
41. K. Hachiya, Y. Ito, Interatomic potentials for rare-earth metals. J. Phys.: Condens. Matter. 11(34), 6543 (1999)
42. B.J. Bucior, N.S. Bobbitt, T. Islamoglu, S. Goswami, A. Gopalan, T. Yildirim, O.K. Farha, N. Bagheri, R.Q. Snurr, Energy-based descriptors to rapidly predict hydrogen storage in metal-organic frameworks. Mol. Syst. Des. Eng. 4(1), 162–174 (2019)10.1039/C8ME00050F ↗
43. E.P. George, D. Raabe, R.O. Ritchie, High-entropy alloys. Nature Rev. Mater. 4(8), 515–534 (2019)10.1038/s41578-019-0121-4 ↗
44. Y. Lysogorskiy, A. Bochkarev, M. Mrovec, R. Drautz, Active learning strategies for atomic cluster expansion models. Phys. Rev. Mater. 7(4), 043801 (2023)10.1103/PhysRevMaterials.7.043801 ↗
45. A.P. Thompson, H.M. Aktulga, R. Berger, D.S. Bolintineanu, W.M. Brown, P.S. Crozier, P.J. Veld, A. Kohlmeyer, S.G. Moore, T.D. Nguyen et al., Lammps-a flexible simulation tool for particle-based materials modeling at the atomic, meso, and continuum scales. Comput. Phys. Commun. 271, 108171 (2022)10.1016/j.cpc.2021.108171 ↗
46. J.S. Smith, N. Lubbers, A.P. Thompson, K. Barros, Simple and efficient algorithms for training machine learning potentials to force data. Preprint at http://arxiv.org/abs/2006.05475 (2020)10.2172/1763572 ↗
47. A. Singraber, T. Morawietz, J. Behler, C. Dellago, Parallel multistream training of high-dimensional neural network potentials. J. Chem. Theory Comput. 15(5), 3075–3092 (2019). https://doi.org/10.1021/acs.jctc.8b01092. (PMID: 30995035)10.1021/acs.jctc.8b01092 ↗