Welcome to the new version of CaltechAUTHORS. Login is currently restricted to library staff. If you notice any issues, please email coda@library.caltech.edu
Published November 2022 | Supplemental Material
Journal Article Open

How to see hidden patterns in metamaterials with interpretable machine learning


Machine learning models can assist with metamaterials design by approximating computationally expensive simulators or solving inverse design problems. However, past work has usually relied on black box deep neural networks, whose reasoning processes are opaque and require enormous datasets that are expensive to obtain. In this work, we develop two novel machine learning approaches to metamaterials discovery that have neither of these disadvantages. These approaches, called shape-frequency features and unit-cell templates, can discover 2D metamaterials with user-specified frequency band gaps. Our approaches provide logical rule-based conditions on metamaterial unit-cells that allow for interpretable reasoning processes, and generalize well across design spaces of different resolutions. The templates also provide design flexibility where users can almost freely design the fine resolution features of a unit-cell without affecting the user's desired band gap.

Additional Information

© 2022 Elsevier. The authors are grateful to M. Bastawrous, A. Lin, K. Liu, C. Zhong, O. Bilal, W. Chen, C. Tomasi, and S. Mukherjee for the feedback and assistance they provided during the develop-ment and preparation of this research. The authors acknowledge funding from the National Science Foundation, United States under grant OAC-1835782, Department of Energy, United States under grants DE-SC0021358 and DE-SC0023194, and National Research Traineeship Program under NSF, United States grants DGE-2022040 and CCF-1934964. CRediT authorship contribution statement. Zhi Chen: Developed the methods, Designed metrics, Designed visualization, Ran experiments related to the ML algorithms, Discussed the results, Contributed to the writing. Alexander Ogren: Developed the simulation code, Generated the data, Did the practicality test, Discussed the results, Contributed to the writing. Chiara Daraio: Conceived, Supervised the project, Discussed the results, Contributed to the writing. L. Catherine Brinson: Conceived, Supervised the project, Discussed the results, Contributed to the writing. Cynthia Rudin: Conceived, Supervised the project, Discussed the results, Contributed to the writing. Code and Data availability. The code and data for replicating our experiments are available on https://github.com/zhiCHEN96/interpretable_ml_metamaterials.git. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Attached Files

Supplemental Material - 1-s2.0-S2352431622001717-mmc1.pdf


Files (7.0 MB)
Name Size Download all
7.0 MB Preview Download

Additional details

August 22, 2023
October 25, 2023