Computational Materials Group

Machine Learning for Materials – Morgan

We are working to integrate machine learning (ML) in materials science and engineering, including for predicting properties and automating microscopy analysis, as well as supporting increased adoption of ML in materials. Some of our major results include:

  1. Development of machine learning for materials property prediction: We have developed ML models for predicting materials properties to aid in generating new materials data and materials discovery. Predicted properties include impurity diffusion2, perovskite compound thermodynamics stability3, electromigration4, concrete mechanical properties5, molecular flash points6, metallic glass critical cooling rates7, irradiation induced hardening of steels8, band gaps 9, and impurity defect levels in semiconductors10. We are also developing uncertainty quantification to make such models more useful, e.g., we recently demonstrated that calibrated ensemble methods could provide quantitative error estimates for model predictions11.
  2. Development of machine vision methods for automatic analysis of defect data in electron microscopy images. In close collaboration with Prof. Kevin Field of University of Michigan we have demonstrated that defects induced by irradiation (e.g., cavities and multiple types of dislocation loops) can be automatically extracted from electron microscopy images using deep learning methods with human level accuracy 12. These methods are expected to provide qualitative change in our ability to characterize and understand defect behavior, in particular their evolution with time and in complex heterogeneous conditions.
  3. Providing critical support for the growing community of materials informatics: Our group and collaborators are strong supporters of the growing field of material informatics. In addition to many presentations in this area we have contributed multiple review papers. These include supporting the overall role of ML in the Materials Genome Initiative 13, a general review of ML in materials with a strong focus on uncertainty quantification1, and a targeted review illustrating how ML might impact nuclear materials 14. In collaboration with Ryan Jacobs we have provided software for the community in the MAterials Simulation Toolkit – Machine Learning (MAST-ML), which has many tools to streamline ML input and output, accelerate basic ML workflows, and support improved uncertainty quantification. We also initiated the Informatics Skunkworks (https://skunkworks.engr.wisc.edu/), a group dedicated to engaging undergraduates in science and engineering informatics research15. The Skunkworks has trained ~400 students since 2015 and provides and open online course and helpdesk for the materials and chemistry communities.
Publication rates on machine learning in materials (updated from Ref.1). (b) Results for a range of property prediction models.

 

 

References

 

1              D. Morgan and R. Jacobs, “Opportunities and Challenges for Machine Learning in Materials Science”, in Annu Rev Mater Res, edited by D. R. Clarke (2020), Vol. 50, pp. 71-103.

2              H. Wu, A. Lorenson, B. Anderson, L. Witteman, H. T. Wu, B. Meredig, and D. Morgan, Robust Fcc Solute Diffusion Predictions from Ab-Initio Machine Learning Methods, Computational Materials Science 134, 160-165 (2017);                 H. J. Lu, N. Zou, R. Jacobs, B. Afflerbach, X. G. Lu, and D. Morgans, Error Assessment and Optimal Cross-Validation Approaches in Machine Learning Applied to Impurity Diffusion, Computational Materials Science 169 (2019).

3              W. Li, R. Jacobs, and D. Morgan, Predicting the Thermodynamic Stability of Perovskite Oxides Using Machine Learning Models, Computational Materials Science 150, 454-463 (2018).

4              Y. C. Liu, B. Afflerbach, R. Jacobs, S. K. Lin, and D. Morgan, Exploring Effective Charge in Electromigration Using Machine Learning, Mrs Communications 9 (2), 567-575 (2019).

5              V. Nilsen, L. T. Pham, M. Hibbard, A. Klager, S. M. Cramer, and D. Morgan, Prediction of Concrete Coefficient of Thermal Expansion and Other Properties Using Machine Learning, Construction and Building Materials 220, 587-595 (2019).

6              X. Y. Sun, N. J. Krakauer, A. Politowicz, W. T. Chen, Q. Y. Li, Z. Y. Li, X. J. Shao, A. Sunaryo, M. R. Shen, J. M. Wang, and D. Morgan, Assessing Graph-Based Deep Learning Models for Predicting Flash Point, Molecular Informatics 39 (6) (2020).

7              B. T. Afflerbach, C. Francis, L. E. Schultz, J. Spethson, V. Meschke, E. Strand, L. Ward, J. H. Perepezko, D. Thoma, P. M. Voyles, I. Szlufarska, and D. Morgan, Machine Learning Prediction of the Critical Cooling Rate for Metallic Glasses from Expanded Datasets and Elemental Features, Chemistry of Materials 34 (7), 2945-2954 (2022).

8            Y. C. Liu, H. Wu, T. Mayeshiba, B. Afflerbach, R. Jacobs, J. Perry, J. George, J. Cordell, J. Y. Xia, H. Yuan, A. Lorenson, H. T. Wu, M. Parker, F. Doshi, A. Politowicz, L. D. Xiao, D. Morgan, P. Wells, N. Almirall, T. Yamamoto, and G. R. Odette, Machine Learning Predictions of Irradiation Embrittlement in Reactor Pressure Vessel Steels, Npj Computational Materials 8 (1) (2022).

9              X. G. Li, B. Blaiszik, M. E. Schwarting, R. Jacobs, A. Scourtas, K. J. Schmidt, P. M. Voyles, and D. Morgan, Graph Network Based Deep Learning of Bandgaps, Journal of Chemical Physics 155 (15) (2021).

10           M. P. Polak, R. Jacobs, A. Mannodi-Kanakkithodi, M. K. Y. Chan, and D. Morgan, Machine Learning for Impurity Charge-State Transition Levels in Semiconductors from Elemental Properties Using Multi-Fidelity Datasets, Journal of Chemical Physics 156 (11) (2022).

11           G. Palmer, S. Q. Du, A. Politowicz, J. P. Emory, X. Y. Yang, A. Gautam, G. Gupta, Z. L. Li, R. Jacobs, and D. Morgan, Calibration after Bootstrap for Accurate Uncertainty Quantification in Regression Models, Npj Computational Materials 8 (1) (2022).

12           W. Li, K. G. Field, and D. Morgan, Automated Defect Analysis in Electron Microscopic Images, Npj Computational Materials 4 (2018);          M. R. Shen, G. Z. Li, D. X. Wu, Y. Yaguchi, J. C. Haley, K. G. Field, and D. Morgan, A Deep Learning Based Automatic Defect Analysis Framework for in-Situ Tem Ion Irradiations, Computational Materials Science 197 (2021);         M. R. Shen, G. Z. Li, D. X. Wu, Y. H. Liu, J. R. C. Greaves, W. Hao, N. J. Krakauer, L. Krudy, J. Perez, V. Sreenivasan, B. Sanchez, O. Torres-Velazquez, W. Li, K. G. Field, and D. Morgan, Multi Defect Detection and Analysis of Electron Microscopy Images with Deep Learning, Computational Materials Science 199 (2021); R. Jacobs, M. R. Shen, Y. H. Liu, W. Hao, X. S. Li, R. Y. He, J. R. C. Greaves, D. L. Wang, Z. M. Xie, Z. T. Huang, C. Wang, K. G. Field, and D. Morgan, Performance and Limitations of Deep Learning Semantic Segmentation of Multiple Defects in Transmission Electron Micrographs, Cell Reports Physical Science 3 (5) (2022).

13           J. J. de Pablo, N. E. Jackson, M. A. Webb, L. Q. Chen, J. E. Moore, D. Morgan, R. Jacobs, T. Pollock, D. G. Schlom, E. S. Toberer, J. Analytis, I. Dabo, D. M. DeLongchamp, G. A. Fiete, G. M. Grason, G. Hautier, Y. F. Mo, K. Rajan, E. J. Reed, E. Rodriguez, V. Stevanovic, J. Suntivich, K. Thornton, and J. C. Zhao, New Frontiers for the Materials Genome Initiative, Npj Computational Materials 5 (2019).

14           D. Morgan, G. Pilania, A. Couet, B. P. Uberuaga, C. Sun, and J. Li, Machine Learning in Nuclear Materials Research, Current Opinion in Solid State & Materials Science 26 (2) (2022).

15           B Afflerbach, N Fathema, A Gillian-Daniel, W Crone, and D. Morgan, Authentic Undergraduate Research in Machine Learning with the Informatics Skunkworks: A Strategy for Scalable Apprenticeship Applied to Materials Informatics Research, 2022 ASEE Annual Conference & Exposition (2022).