pith. sign in

arxiv: 2401.00003 · v6 · submitted 2023-12-08 · ⚛️ physics.optics · cs.LG

Generative Inverse Design of Metamaterials with Functional Responses by Interpretable Learning

Wei "Wayne" Chen , Rachel Sun , Doksoo Lee , Carlos M. Portela , Wei Chen This is my paper

Pith reviewed 2026-05-24 05:32 UTC · model grok-4.3

classification ⚛️ physics.optics cs.LG
keywords designinversefunctionalrigidgenerativeinterpretablemetamaterialsresponses
0
0 comments X

The pith

RIGID uses a random forest forward model and MCMC sampling to generate metamaterial designs satisfying target functional responses, producing broader design-space coverage than genetic algorithms on acoustic and optical test cases with fewer than 250 training samples.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Metamaterials are engineered structures whose properties like wave response or stiffness can change with conditions. Finding designs that hit specific target behaviors is hard because many different shapes can produce similar responses and data is often scarce. The paper proposes RIGID, which first builds a random forest model that predicts the response from a given design. Because random forests can output probabilities, the method turns those probabilities into a likelihood score for how well a design meets the target. It then runs Markov chain Monte Carlo sampling to draw many candidate designs that score high on that likelihood. This avoids training a separate inverse network that maps responses back to designs. The approach is tested on two small-data problems in acoustics and optics. Compared with a genetic algorithm baseline, the sampled designs cover more of the possible design space while still meeting the targets, which leaves room to optimize other criteria such as manufacturability.

Core claim

RIGID generates satisfactory solutions that cover a broader range of the design space, allowing for better consideration of additional figures of merit beyond target satisfaction.

Load-bearing premise

The random forest trained on the forward mapping supplies sufficiently accurate likelihood estimates that MCMC sampling from those likelihoods reliably produces valid designs satisfying the qualitative functional targets.

Figures

Figures reproduced from arXiv: 2401.00003 by Carlos M. Portela, Doksoo Lee, Rachel Sun, Wei Chen, Wei "Wayne" Chen.

Figure 1
Figure 1. Figure 1: Schematic diagram of the RIGID method. We first train a random forest on a design-response dataset to learn [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The inverse design pipeline of the proposed RIGID method (using the inverse design of acoustic metamaterials [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Acoustic metamaterial design problem configuration and results. (A) Design variables of center and corner [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Distributions of satisfactory solutions for two bandgap targets. The off-diagonal plots show the pairwise bivariate distributions of design variables, and the diagonal plots show the marginal distributions of the data in each column. The left panel shows that GA designs are highly localized while RIGID can lead to diverse solutions. The right panel indicates that none of the GA designs satisfy the target, … view at source ↗
Figure 5
Figure 5. Figure 5: Optical metasurface design problem configuration and results. (A-B) Design variables (materials, layer [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Synthetic data creation for (A) the SqExp problem and (B) the SupSin problem. For each problem, the left [PITH_FULL_IMAGE:figures/full_fig_p012_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Visualization of estimated likelihood and validation metrics for synthetic problems. (A) Likelihood function [PITH_FULL_IMAGE:figures/full_fig_p013_7.png] view at source ↗
read the original abstract

Metamaterials with functional responses can exhibit varying properties under different conditions (e.g., wave-based responses or deformation-induced property variation). This work addresses the rapid inverse design of such metamaterials to meet target qualitative functional behaviors, a challenge due to its intractability and non-unique solutions. Unlike data-intensive and non-interpretable deep-learning-based methods, we propose the Random-forest-based Interpretable Generative Inverse Design (RIGID), a single-shot inverse design method for fast generation of metamaterial designs with on-demand functional behaviors. RIGID leverages the interpretability of a random forest-based "design$\rightarrow$response" forward model, eliminating the need for a more complex "response$\rightarrow$design" inverse model. Based on the likelihood of target satisfaction derived from the trained random forest, one can sample a desired number of design solutions using Markov chain Monte Carlo methods. We validate RIGID on acoustic and optical metamaterial design problems, each with fewer than 250 training samples. Compared to the genetic algorithm-based design generation approach, RIGID generates satisfactory solutions that cover a broader range of the design space, allowing for better consideration of additional figures of merit beyond target satisfaction. This work offers a new perspective on solving on-demand inverse design problems, showcasing the potential for incorporating interpretable machine learning into generative design under small data constraints.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript proposes RIGID, a random-forest-based method for single-shot inverse design of metamaterials exhibiting qualitative functional responses. A forward RF model is trained on design-to-response data (<250 samples for acoustic and optical examples), likelihoods of target satisfaction are extracted, and MCMC is used to sample candidate designs. The central claim is that the resulting solutions are satisfactory while covering a broader region of design space than a genetic-algorithm baseline, thereby permitting additional figures of merit to be considered.

Significance. If the calibration and coverage claims hold, the approach supplies an interpretable, data-efficient generative alternative to deep-learning inverse models for functional metamaterial design, directly addressing the small-data regime common in metamaterials research.

major comments (3)
  1. [Method and Validation] The central claim that RIGID produces valid designs via MCMC driven by RF-derived likelihoods rests on the assumption that those likelihoods are sufficiently well-calibrated. With training sets smaller than 250 points and qualitative (potentially discontinuous) targets, random-forest probability estimates are known to be poorly calibrated; the manuscript must demonstrate, via held-out simulator validation or calibration plots, that the MCMC samples actually satisfy the targets at rates consistent with the reported likelihoods.
  2. [Results] The assertion of 'broader design-space coverage' relative to GA is load-bearing for the practical advantage claimed. The manuscript must supply quantitative metrics (e.g., convex-hull volume in normalized parameter space, average pairwise distance, or coverage of a reference grid) together with statistical error bars; a qualitative statement is insufficient to support the conclusion that additional figures of merit can be reliably traded off.
  3. [Results] The comparison to GA does not specify whether the GA was run with equivalent computational budget, population size, or convergence criteria, nor whether the same forward simulator was used for both methods. Without these controls, the reported superiority in coverage cannot be attributed to the RF+MCMC procedure rather than implementation differences.
minor comments (2)
  1. [Method] Notation for the likelihood function derived from the random-forest ensemble should be defined explicitly (e.g., how leaf probabilities are aggregated across trees).
  2. [Figures] Figure captions should state the exact number of training samples, the number of MCMC samples retained, and the acceptance rate for each example.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive comments. We address each major point below and have revised the manuscript to strengthen the validation and comparisons.

read point-by-point responses

  1. Referee: [Method and Validation] The central claim that RIGID produces valid designs via MCMC driven by RF-derived likelihoods rests on the assumption that those likelihoods are sufficiently well-calibrated. With training sets smaller than 250 points and qualitative (potentially discontinuous) targets, random-forest probability estimates are known to be poorly calibrated; the manuscript must demonstrate, via held-out simulator validation or calibration plots, that the MCMC samples actually satisfy the targets at rates consistent with the reported likelihoods.

    Authors: We agree that explicit calibration validation is necessary given the small training sets and qualitative targets. The original manuscript did not include held-out simulator checks or calibration plots for the RF likelihoods. In the revised version we have added reliability diagrams and results from evaluating MCMC samples with the full simulator, confirming that observed target satisfaction rates align with the reported likelihoods for both test cases. revision: yes

  2. Referee: [Results] The assertion of 'broader design-space coverage' relative to GA is load-bearing for the practical advantage claimed. The manuscript must supply quantitative metrics (e.g., convex-hull volume in normalized parameter space, average pairwise distance, or coverage of a reference grid) together with statistical error bars; a qualitative statement is insufficient to support the conclusion that additional figures of merit can be reliably traded off.

    Authors: We concur that quantitative metrics with error bars are required. The original manuscript supported broader coverage only through visualizations. The revision now reports convex-hull volumes in normalized design space and mean pairwise distances, each accompanied by standard errors from multiple independent runs, providing statistical support for the coverage advantage. revision: yes

  3. Referee: [Results] The comparison to GA does not specify whether the GA was run with equivalent computational budget, population size, or convergence criteria, nor whether the same forward simulator was used for both methods. Without these controls, the reported superiority in coverage cannot be attributed to the RF+MCMC procedure rather than implementation differences.

    Authors: We thank the referee for noting the missing controls. The same forward simulator was used for both methods, but the manuscript did not detail GA hyperparameters or budget equivalence. The revision now specifies the GA population size, generation count, convergence criteria, and confirms that total simulator evaluations were matched, allowing the coverage difference to be attributed to the RF+MCMC approach. revision: yes

Circularity Check

0 steps flagged

No circularity; standard surrogate + MCMC pipeline

full rationale

The paper trains a random forest forward model on design-to-response pairs, derives likelihoods of target satisfaction from that model, and samples designs via MCMC. This is a conventional surrogate-assisted generative procedure with no equations or claims that reduce a prediction to a fitted quantity defined by the same data, no load-bearing self-citations, and no ansatz or uniqueness result imported from prior author work. The comparison to genetic algorithms is external and falsifiable. The derivation chain is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review yields no explicit free parameters, axioms, or invented entities; the approach rests on the standard assumption that random forests can be interpreted as probabilistic forward models.

pith-pipeline@v0.9.0 · 5785 in / 1128 out tokens · 23172 ms · 2026-05-24T05:32:36.028014+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

98 extracted references · 98 canonical work pages · 1 internal anchor

  1. [1]

    Metamaterials beyond electromagnetism

    Muamer Kadic, Tiemo Bückmann, Robert Schittny, and Martin Wegener. Metamaterials beyond electromagnetism. Reports on Progress in physics, 76(12):126501, 2013

    work page 2013

  2. [2]

    Flexible mechanical metamaterials

    Katia Bertoldi, Vincenzo Vitelli, Johan Christensen, and Martin Van Hecke. Flexible mechanical metamaterials. Nature Reviews Materials, 2(11):1–11, 2017. 16

    work page 2017

  3. [3]

    Past achievements and future challenges in the development of three-dimensional photonic metamaterials

    Costas M Soukoulis and Martin Wegener. Past achievements and future challenges in the development of three-dimensional photonic metamaterials. Nature photonics, 5(9):523–530, 2011

    work page 2011

  4. [4]

    Controlling sound with acoustic metamaterials

    Steven A Cummer, Johan Christensen, and Andrea Alù. Controlling sound with acoustic metamaterials. Nature Reviews Materials, 1(3):1–13, 2016

    work page 2016

  5. [5]

    Experiments on transformation thermodynamics: molding the flow of heat

    Robert Schittny, Muamer Kadic, Sebastien Guenneau, and Martin Wegener. Experiments on transformation thermodynamics: molding the flow of heat. Physical review letters, 110(19):195901, 2013

    work page 2013

  6. [6]

    Vibrant times for mechanical metamaterials

    Johan Christensen, Muamer Kadic, Oliver Kraft, and Martin Wegener. Vibrant times for mechanical metamaterials. Mrs Communications, 5(3):453–462, 2015

    work page 2015

  7. [7]

    Negative refraction makes a perfect lens

    John Brian Pendry. Negative refraction makes a perfect lens. Physical review letters, 85(18):3966, 2000

    work page 2000

  8. [8]

    Acoustic metamaterials for noise reduction: a review

    Nansha Gao, Zhicheng Zhang, Jie Deng, Xinyu Guo, Baozhu Cheng, and Hong Hou. Acoustic metamaterials for noise reduction: a review. Advanced Materials Technologies, 7(6):2100698, 2022

    work page 2022

  9. [9]

    Flexible metamaterial electronics

    Shan Jiang, Xuejun Liu, Jianpeng Liu, Dong Ye, Yongqing Duan, Kan Li, Zhouping Yin, and YongAn Huang. Flexible metamaterial electronics. Advanced Materials, 34(52):2200070, 2022

    work page 2022

  10. [10]

    Shape-shifting structured lattices via multimaterial 4d printing

    J William Boley, Wim M Van Rees, Charles Lissandrello, Mark N Horenstein, Ryan L Truby, Arda Kotikian, Jennifer A Lewis, and L Mahadevan. Shape-shifting structured lattices via multimaterial 4d printing. Proceedings of the National Academy of Sciences, 116(42):20856–20862, 2019

    work page 2019

  11. [11]

    Deep learning-accelerated designs of tunable magneto-mechanical metamaterials

    Chunping Ma, Yilong Chang, Shuai Wu, and Ruike Renee Zhao. Deep learning-accelerated designs of tunable magneto-mechanical metamaterials. ACS Applied Materials & Interfaces, 14(29):33892–33902, 2022

    work page 2022

  12. [12]

    Piezoelectric resonator arrays for tunable acoustic waveguides and metamaterials

    Filippo Casadei, Tommaso Delpero, Andrea Bergamini, Paolo Ermanni, and Massimo Ruzzene. Piezoelectric resonator arrays for tunable acoustic waveguides and metamaterials. Journal of Applied Physics, 112(6):064902, 09 2012

    work page 2012

  13. [13]

    Acoustic waveguide with virtual soft boundary based on metamaterials

    Guang-Sheng Liu, Yang Zhou, Ming-Hao Liu, Ying Yuan, Xin-Ye Zou, and Jian-Chun Cheng. Acoustic waveguide with virtual soft boundary based on metamaterials. Scientific Reports, 10(1):981, 2020

    work page 2020

  14. [14]

    Poroelastic microlattices for underwater wave focusing

    Gunho Kim, Carlos M Portela, Paolo Celli, Antonio Palermo, and Chiara Daraio. Poroelastic microlattices for underwater wave focusing. Extreme Mechanics Letters, 49:101499, 2021

    work page 2021

  15. [15]

    Acoustic imaging with metamaterial luneburg lenses

    Yangbo Xie, Yangyang Fu, Zhetao Jia, Junfei Li, Chen Shen, Yadong Xu, Huanyang Chen, and Steven A Cummer. Acoustic imaging with metamaterial luneburg lenses. Scientific reports, 8(1):16188, 2018

    work page 2018

  16. [16]

    Composite 3d-printed metastructures for low-frequency and broadband vibration absorption

    Kathryn H Matlack, Anton Bauhofer, Sebastian Krödel, Antonio Palermo, and Chiara Daraio. Composite 3d-printed metastructures for low-frequency and broadband vibration absorption. Proceedings of the National Academy of Sciences, 113(30):8386–8390, 2016

    work page 2016

  17. [17]

    Wide band-gap seismic metastructures

    Sebastian Krödel, Nicolas Thomé, and Chiara Daraio. Wide band-gap seismic metastructures. Extreme Mechanics Letters, 4:111–117, 2015

    work page 2015

  18. [18]

    Wave directionality in three-dimensional periodic lattices

    Alireza Bayat and Stavros Gaitanaros. Wave directionality in three-dimensional periodic lattices. Journal of Applied Mechanics, 85(1):011004, 2018

    work page 2018

  19. [19]

    Inverse design of discrete mechanical metamaterials

    Henrik Ronellenfitsch, Norbert Stoop, Josephine Yu, Aden Forrow, and Jörn Dunkel. Inverse design of discrete mechanical metamaterials. Physical Review Materials, 3(9):095201, 2019

    work page 2019

  20. [20]

    Inverse band gap design of elastic metamaterials for p and sv wave control

    Heedong Goh and Loukas F Kallivokas. Inverse band gap design of elastic metamaterials for p and sv wave control. Computer Methods in Applied Mechanics and Engineering, 370:113263, 2020

    work page 2020

  21. [21]

    Designing phononic crystal with anticipated band gap through a deep learning based data-driven method

    Xiang Li, Shaowu Ning, Zhanli Liu, Ziming Yan, Chengcheng Luo, and Zhuo Zhuang. Designing phononic crystal with anticipated band gap through a deep learning based data-driven method. Computer Methods in Applied Mechanics and Engineering, 361:112737, 2020

    work page 2020

  22. [22]

    Realization of full and directional band gap design by non-gradient topology optimization in acoustic metamaterials

    Xiaopeng Zhang, Jian Xing, Pai Liu, Yangjun Luo, and Zhan Kang. Realization of full and directional band gap design by non-gradient topology optimization in acoustic metamaterials. Extreme Mechanics Letters, 42:101126, 2021

    work page 2021

  23. [23]

    How to see hidden patterns in metamaterials with interpretable machine learning

    Zhi Chen, Alexander Ogren, Chiara Daraio, L Catherine Brinson, and Cynthia Rudin. How to see hidden patterns in metamaterials with interpretable machine learning. Extreme Mechanics Letters, 57:101895, 2022

    work page 2022

  24. [24]

    Solar harvesting based on perfect absorbing all-dielectric nanoresonators on a mirror

    Robin Vismara, Nils Odebo Länk, Ruggero Verre, Mikael Käll, Olindo Isabella, and Miro Zeman. Solar harvesting based on perfect absorbing all-dielectric nanoresonators on a mirror. Optics Express, 27(16):A967–A980, 2019

    work page 2019

  25. [25]

    Plasmonic chiral metasurface absorber based on bilayer fourfold twisted semicircle nanostructure at optical frequency

    Yongzhi Cheng, Fu Chen, and Hui Luo. Plasmonic chiral metasurface absorber based on bilayer fourfold twisted semicircle nanostructure at optical frequency. Nanoscale Research Letters, 16:1–9, 2021. 17

    work page 2021

  26. [26]

    Compounding meta-atoms into metamolecules with hybrid artificial intelligence techniques

    Zhaocheng Liu, Dayu Zhu, Kyu-Tae Lee, Andrew S Kim, Lakshmi Raju, and Wenshan Cai. Compounding meta-atoms into metamolecules with hybrid artificial intelligence techniques. Advanced Materials, 32(6):1904790, 2020

    work page 2020

  27. [27]

    Deep neural operator enabled concurrent multitask design for multifunctional metamaterials under heterogeneous fields

    Doksoo Lee, Lu Zhang, Yue Yu, and Wei Chen. Deep neural operator enabled concurrent multitask design for multifunctional metamaterials under heterogeneous fields. Advanced Optical Materials, 12(15):2303087, 2024

    work page 2024

  28. [28]

    Neural network based design of metagratings

    Sandeep Inampudi and Hossein Mosallaei. Neural network based design of metagratings. Applied Physics Letters, 112(24), 2018

    work page 2018

  29. [29]

    Deep learning to accelerate scatterer-to-field mapping for inverse design of dielectric metasurfaces

    Maksym V Zhelyeznyakov, Steve Brunton, and Arka Majumdar. Deep learning to accelerate scatterer-to-field mapping for inverse design of dielectric metasurfaces. ACS Photonics, 8(2):481–488, 2021

    work page 2021

  30. [30]

    Training deep neural networks for the inverse design of nanophotonic structures

    Dianjing Liu, Yixuan Tan, Erfan Khoram, and Zongfu Yu. Training deep neural networks for the inverse design of nanophotonic structures. Acs Photonics, 5(4):1365–1369, 2018

    work page 2018

  31. [31]

    Probabilistic representation and inverse design of metamaterials based on a deep generative model with semi-supervised learning strategy

    Wei Ma, Feng Cheng, Yihao Xu, Qinlong Wen, and Yongmin Liu. Probabilistic representation and inverse design of metamaterials based on a deep generative model with semi-supervised learning strategy. Advanced Materials, 31(35):1901111, 2019

    work page 2019

  32. [32]

    Shalaginov, Li Zhou, Hang Li, Myungkoo Kang, Kathleen A

    Sensong An, Bowen Zheng, Hong Tang, Mikhail Y . Shalaginov, Li Zhou, Hang Li, Myungkoo Kang, Kathleen A. Richardson, Tian Gu, Juejun Hu, Clayton Fowler, and Hualiang Zhang. Multifunctional Metasurface Design with a Generative Adversarial Network. Advanced Optical Materials, 9(5):1–10, 2021

    work page 2021

  33. [33]

    Plasmonic nanostruc- ture design and characterization via Deep Learning

    Itzik Malkiel, Michael Mrejen, Achiya Nagler, Uri Arieli, Lior Wolf, and Haim Suchowski. Plasmonic nanostruc- ture design and characterization via Deep Learning. Light: Science and Applications, 7(1), 2018

    work page 2018

  34. [34]

    Data-driven design for metamaterials and multiscale systems: A review

    Doksoo Lee, Wei Chen, Liwei Wang, Yu-Chin Chan, and Wei Chen. Data-driven design for metamaterials and multiscale systems: A review. Advanced Materials, 36(8):2305254, 2024

    work page 2024

  35. [35]

    A deep learning approach for objective-driven all-dielectric metasurface design

    Sensong An, Clayton Fowler, Bowen Zheng, Mikhail Y Shalaginov, Hong Tang, Hang Li, Li Zhou, Jun Ding, Anuradha Murthy Agarwal, Clara Rivero-Baleine, et al. A deep learning approach for objective-driven all-dielectric metasurface design. ACS Photonics, 6(12):3196–3207, 2019

    work page 2019

  36. [36]

    Inverse-designed spinodoid metamaterials

    Siddhant Kumar, Stephanie Tan, Li Zheng, and Dennis M Kochmann. Inverse-designed spinodoid metamaterials. npj Computational Materials, 6(1):1–10, 2020

    work page 2020

  37. [37]

    Knight, and Aaswath P

    Christopher Yeung, Ju Ming Tsai, Brian King, Benjamin Pham, David Ho, Julia Liang, Mark W. Knight, and Aaswath P. Raman. Multiplexed supercell metasurface design and optimization with tandem residual networks. Nanophotonics, 10(3):1133–1143, 1 2021

    work page 2021

  38. [38]

    Inverting the structure–property map of truss metamaterials by deep learning

    Jan-Hendrik Bastek, Siddhant Kumar, Bastian Telgen, Raphaël N Glaesener, and Dennis M Kochmann. Inverting the structure–property map of truss metamaterials by deep learning. Proceedings of the National Academy of Sciences, 119(1):e2111505119, 2022

    work page 2022

  39. [39]

    Deep learning assisted inverse design of metamaterial microwave absorber

    Chen Xie, Haonan Li, Chenyang Cui, Haodong Lei, Yingjie Sun, Chi Zhang, Yaqiang Zhang, Hongxing Dong, and Long Zhang. Deep learning assisted inverse design of metamaterial microwave absorber. Applied Physics Letters, 123(18), 2023

    work page 2023

  40. [40]

    A deep autoencoder based approach for the inverse design of an acoustic-absorber

    K Mahesh, S Kumar Ranjith, and RS Mini. A deep autoencoder based approach for the inverse design of an acoustic-absorber. Engineering with Computers, 40(1):279–300, 2024

    work page 2024

  41. [41]

    Rapid inverse design of metamaterials based on prescribed mechanical behavior through machine learning

    Chan Soo Ha, Desheng Yao, Zhenpeng Xu, Chenang Liu, Han Liu, Daniel Elkins, Matthew Kile, Vikram Deshpande, Zhenyu Kong, Mathieu Bauchy, et al. Rapid inverse design of metamaterials based on prescribed mechanical behavior through machine learning. Nature Communications, 14(1):5765, 2023

    work page 2023

  42. [42]

    On-demand inverse design of acoustic metamaterials using probabilistic generation network

    Ze-Wei Wang, An Chen, Zi-Xiang Xu, Jing Yang, Bin Liang, and Jian-Chun Cheng. On-demand inverse design of acoustic metamaterials using probabilistic generation network. Science China Physics, Mechanics & Astronomy, 66(2):224311, 2023

    work page 2023

  43. [43]

    Free-form diffractive metagrating design based on generative adversarial networks

    Jiaqi Jiang, David Sell, Stephan Hoyer, Jason Hickey, Jianji Yang, and Jonathan A Fan. Free-form diffractive metagrating design based on generative adversarial networks. ACS nano, 13(8):8872–8878, 2019

    work page 2019

  44. [44]

    Designing nanophotonic structures using conditional deep convolutional generative adversarial networks

    Sunae So and Junsuk Rho. Designing nanophotonic structures using conditional deep convolutional generative adversarial networks. Nanophotonics, 8(7):1255–1261, 2019

    work page 2019

  45. [45]

    Robust freeform metasurface design based on progressively growing generative networks

    Fufang Wen, Jiaqi Jiang, and Jonathan A Fan. Robust freeform metasurface design based on progressively growing generative networks. ACS Photonics, 7(8):2098–2104, 2020

    work page 2098

  46. [46]

    Generative adversarial networks for the design of acoustic metamaterials

    Caglar Gurbuz, Felix Kronowetter, Christoph Dietz, Martin Eser, Jonas Schmid, and Steffen Marburg. Generative adversarial networks for the design of acoustic metamaterials. The Journal of the Acoustical Society of America, 149(2):1162–1174, 2021. 18

    work page 2021

  47. [47]

    Pushing the limits of functionality-multiplexing capability in metasurface design based on statistical machine learning

    Wei Ma, Yihao Xu, Bo Xiong, Lin Deng, Ru-Wen Peng, Mu Wang, and Yongmin Liu. Pushing the limits of functionality-multiplexing capability in metasurface design based on statistical machine learning. Advanced Materials, 34(16):2110022, 2022

    work page 2022

  48. [48]

    Dynamic multifunctional metasurfaces: an inverse design deep learning approach

    Zhi-Dan Lei, Yi-Duo Xu, Cheng Lei, Yan Zhao, and Du Wang. Dynamic multifunctional metasurfaces: an inverse design deep learning approach. Photonics Research, 12(1):123–133, 2024

    work page 2024

  49. [49]

    Single-shot forward and inverse hierarchical architected materials design for nonlinear mechanical properties using an attention-diffusion model

    Andrew J Lew and Markus J Buehler. Single-shot forward and inverse hierarchical architected materials design for nonlinear mechanical properties using an attention-diffusion model. Materials Today, 64:10–20, 2023

    work page 2023

  50. [50]

    Diffusion probabilistic model based accurate and high-degree-of-freedom metasurface inverse design.Nanophotonics, 12(20):3871–3881, 2023

    Zezhou Zhang, Chuanchuan Yang, Yifeng Qin, Hao Feng, Jiqiang Feng, and Hongbin Li. Diffusion probabilistic model based accurate and high-degree-of-freedom metasurface inverse design.Nanophotonics, 12(20):3871–3881, 2023

    work page 2023

  51. [51]

    Inverse design of nonlinear mechanical metamaterials via video denoising diffusion models

    Jan-Hendrik Bastek and Dennis M Kochmann. Inverse design of nonlinear mechanical metamaterials via video denoising diffusion models. Nature Machine Intelligence, 5(12):1466–1475, 2023

    work page 2023

  52. [52]

    Revisiting the design strategies for metasurfaces: fundamental physics, optimization, and beyond

    Sunae So, Jungho Mun, Junghyun Park, and Junsuk Rho. Revisiting the design strategies for metasurfaces: fundamental physics, optimization, and beyond. Advanced Materials, 35(43):2206399, 2023

    work page 2023

  53. [53]

    Deep learning in mechanical metamaterials: from prediction and generation to inverse design

    Xiaoyang Zheng, Xubo Zhang, Ta-Te Chen, and Ikumu Watanabe. Deep learning in mechanical metamaterials: from prediction and generation to inverse design. Advanced Materials, 35(45):2302530, 2023

    work page 2023

  54. [54]

    Intelligent on-demand design of phononic metamaterials

    Yabin Jin, Liangshu He, Zhihui Wen, Bohayra Mortazavi, Hongwei Guo, Daniel Torrent, Bahram Djafari-Rouhani, Timon Rabczuk, Xiaoying Zhuang, and Yan Li. Intelligent on-demand design of phononic metamaterials. Nanophotonics, 11(3):439–460, 2022

    work page 2022

  55. [55]

    Deep learning for the design of photonic structures

    Wei Ma, Zhaocheng Liu, Zhaxylyk A Kudyshev, Alexandra Boltasseva, Wenshan Cai, and Yongmin Liu. Deep learning for the design of photonic structures. Nature Photonics, 15(2):77–90, 2021

    work page 2021

  56. [56]

    Deep neural networks for the evaluation and design of photonic devices

    Jiaqi Jiang, Mingkun Chen, and Jonathan A Fan. Deep neural networks for the evaluation and design of photonic devices. Nature Reviews Materials, 6(8):679–700, 2021

    work page 2021

  57. [57]

    Software-defined nanophotonic devices and systems empowered by machine learning

    Yihao Xu, Bo Xiong, Wei Ma, and Yongmin Liu. Software-defined nanophotonic devices and systems empowered by machine learning. Progress in Quantum Electronics, 89:100469, 2023

    work page 2023

  58. [58]

    Interpretable forward and inverse design of particle spectral emissivity using common machine-learning models

    Mahmoud Elzouka, Charles Yang, Adrian Albert, Ravi S Prasher, and Sean D Lubner. Interpretable forward and inverse design of particle spectral emissivity using common machine-learning models. Cell Reports Physical Science, 1(12), 2020

    work page 2020

  59. [59]

    Random forests

    Leo Breiman. Random forests. Machine learning, 45:5–32, 2001

    work page 2001

  60. [60]

    W. K. Hastings. Monte Carlo sampling methods using Markov chains and their applications. Biometrika, 57(1):97–109, 1970

    work page 1970

  61. [61]

    Classification and Regression Trees

    Leo Breiman, Jerome Friedman, Charles J Stone, and R A Olshen. Classification and Regression Trees. Chapman and Hall/CRC, 1984

    work page 1984

  62. [62]

    Smote: synthetic minority over-sampling technique

    Nitesh V Chawla, Kevin W Bowyer, Lawrence O Hall, and W Philip Kegelmeyer. Smote: synthetic minority over-sampling technique. Journal of artificial intelligence research, 16:321–357, 2002

    work page 2002

  63. [63]

    A hybrid strategy for the discovery and design of photonic structures

    Zhaocheng Liu, Lakshmi Raju, Dayu Zhu, and Wenshan Cai. A hybrid strategy for the discovery and design of photonic structures. IEEE Journal on Emerging and Selected Topics in Circuits and Systems, 10(1):126–135, 2020

    work page 2020

  64. [64]

    Pragmatic generative optimization of novel structural lattice metamaterials with machine learning

    Anthony P Garland, Benjamin C White, Scott C Jensen, and Brad L Boyce. Pragmatic generative optimization of novel structural lattice metamaterials with machine learning. Materials & Design, 203:109632, 2021

    work page 2021

  65. [65]

    Genetic algorithms to automate the design of metasurfaces for absorption bandwidth broadening

    Junming Zhang, Guowu Wang, Tao Wang, and Fashen Li. Genetic algorithms to automate the design of metasurfaces for absorption bandwidth broadening. ACS Applied Materials & Interfaces, 13(6):7792–7800, 2021

    work page 2021

  66. [66]

    Nature-inspired architected materials using unsupervised deep learning

    Sabrina Chin-yun Shen and Markus J Buehler. Nature-inspired architected materials using unsupervised deep learning. Communications Engineering, 1(1):37, 2022

    work page 2022

  67. [67]

    Generative machine learning algorithm for lattice structures with superior mechanical properties

    Sangryun Lee, Zhizhou Zhang, and Grace X Gu. Generative machine learning algorithm for lattice structures with superior mechanical properties. Materials Horizons, 9(3):952–960, 2022

    work page 2022

  68. [68]

    An introduction to genetic algorithms

    Melanie Mitchell. An introduction to genetic algorithms. MIT press, 1998

    work page 1998

  69. [69]

    Light propagation with phase discontinuities: generalized laws of reflection and refraction

    Nanfang Yu, Patrice Genevet, Mikhail A Kats, Francesco Aieta, Jean-Philippe Tetienne, Federico Capasso, and Zeno Gaburro. Light propagation with phase discontinuities: generalized laws of reflection and refraction. science, 334(6054):333–337, 2011. 19

    work page 2011

  70. [70]

    Planar photonics with metasurfaces

    Alexander V Kildishev, Alexandra Boltasseva, and Vladimir M Shalaev. Planar photonics with metasurfaces. Science, 339(6125):1232009, 2013

    work page 2013

  71. [71]

    Coding metamaterials, digital metamaterials and programmable metamaterials

    Tie Jun Cui, Mei Qing Qi, Xiang Wan, Jie Zhao, and Qiang Cheng. Coding metamaterials, digital metamaterials and programmable metamaterials. Light: science & applications, 3(10):e218–e218, 2014

    work page 2014

  72. [72]

    A metasurfaces review: Definitions and applications

    Syed S Bukhari, J Yiannis Vardaxoglou, and William Whittow. A metasurfaces review: Definitions and applications. Applied Sciences, 9(13):2727, 2019

    work page 2019

  73. [73]

    A review on metasurface: from principle to smart metadevices

    Jie Hu, Sankhyabrata Bandyopadhyay, Yu-hui Liu, and Li-yang Shao. A review on metasurface: from principle to smart metadevices. Frontiers in Physics, 8:586087, 2021

    work page 2021

  74. [74]

    Perfect metamaterial absorber

    N Iê Landy, S Sajuyigbe, Jack J Mock, David R Smith, and Willie J Padilla. Perfect metamaterial absorber. Physical review letters, 100(20):207402, 2008

    work page 2008

  75. [75]

    Infrared spatial and frequency selective metamaterial with near-unity absorbance

    Xianliang Liu, Tatiana Starr, Anthony F Starr, and Willie J Padilla. Infrared spatial and frequency selective metamaterial with near-unity absorbance. Physical review letters, 104(20):207403, 2010

    work page 2010

  76. [76]

    High performance optical absorber based on a plasmonic metamaterial

    Jiaming Hao, Jing Wang, Xianliang Liu, Willie J Padilla, Lei Zhou, and Min Qiu. High performance optical absorber based on a plasmonic metamaterial. Applied Physics Letters, 96(25), 2010

    work page 2010

  77. [77]

    Metamaterial electromagnetic wave absorbers

    Claire M Watts, Xianliang Liu, and Willie J Padilla. Metamaterial electromagnetic wave absorbers. Advanced materials, 24(23):OP98–OP120, 2012

    work page 2012

  78. [78]

    Plasmonic and metamaterial structures as electromagnetic absorbers

    Yanxia Cui, Yingran He, Yi Jin, Fei Ding, Liu Yang, Yuqian Ye, Shoumin Zhong, Yinyue Lin, and Sailing He. Plasmonic and metamaterial structures as electromagnetic absorbers. Laser & Photonics Reviews, 8(4):495–520, 2014

    work page 2014

  79. [79]

    Experimental realization of a terahertz all-dielectric metasurface absorber

    Xinyu Liu, Kebin Fan, Ilya V Shadrivov, and Willie J Padilla. Experimental realization of a terahertz all-dielectric metasurface absorber. Optics express, 25(1):191–201, 2017

    work page 2017

  80. [80]

    Complete amplitude and phase control of light using broadband holographic metasurfaces

    Gun-Yeal Lee, Gwanho Yoon, Seung-Yeol Lee, Hansik Yun, Jaebum Cho, Kyookeun Lee, Hwi Kim, Junsuk Rho, and Byoungho Lee. Complete amplitude and phase control of light using broadband holographic metasurfaces. Nanoscale, 10(9):4237–4245, 2018

    work page 2018

Showing first 80 references.