Skip to main content
Journal cover image

Transfer learning for metamaterial design and simulation

Publication ,  Journal Article
Peng, R; Ren, S; Malof, J; Padilla, WJ
Published in: Nanophotonics
May 3, 2024

We demonstrate transfer learning as a tool to improve the efficacy of training deep learning models based on residual neural networks (ResNets). Specifically, we examine its use for study of multi-scale electrically large metasurface arrays under open boundary conditions in electromagnetic metamaterials. Our aim is to assess the efficiency of transfer learning across a range of problem domains that vary in their resemblance to the original base problem for which the ResNet model was initially trained. We use a quasi-analytical discrete dipole approximation (DDA) method to simulate electrically large metasurface arrays to obtain ground truth data for training and testing of our deep neural network. Our approach can save significant time for examining novel metasurface designs by harnessing the power of transfer learning, as it effectively mitigates the pervasive data bottleneck issue commonly encountered in deep learning. We demonstrate that for the best case when the transfer task is sufficiently similar to the target task, a new task can be effectively trained using only a few data points yet still achieve a test mean absolute relative error of 3% with a pre-trained neural network, realizing data reduction by a factor of 1000.

Duke Scholars

Published In

Nanophotonics

DOI

EISSN

2192-8614

Publication Date

May 3, 2024

Volume

13

Issue

13

Start / End Page

2323 / 2334

Related Subject Headings

  • 5108 Quantum physics
  • 5102 Atomic, molecular and optical physics
  • 4018 Nanotechnology
  • 1007 Nanotechnology
  • 0906 Electrical and Electronic Engineering
  • 0205 Optical Physics
 

Citation

APA
Chicago
ICMJE
MLA
NLM
Peng, R., Ren, S., Malof, J., & Padilla, W. J. (2024). Transfer learning for metamaterial design and simulation. Nanophotonics, 13(13), 2323–2334. https://doi.org/10.1515/nanoph-2023-0691
Peng, R., S. Ren, J. Malof, and W. J. Padilla. “Transfer learning for metamaterial design and simulation.” Nanophotonics 13, no. 13 (May 3, 2024): 2323–34. https://doi.org/10.1515/nanoph-2023-0691.
Peng R, Ren S, Malof J, Padilla WJ. Transfer learning for metamaterial design and simulation. Nanophotonics. 2024 May 3;13(13):2323–34.
Peng, R., et al. “Transfer learning for metamaterial design and simulation.” Nanophotonics, vol. 13, no. 13, May 2024, pp. 2323–34. Scopus, doi:10.1515/nanoph-2023-0691.
Peng R, Ren S, Malof J, Padilla WJ. Transfer learning for metamaterial design and simulation. Nanophotonics. 2024 May 3;13(13):2323–2334.
Journal cover image

Published In

Nanophotonics

DOI

EISSN

2192-8614

Publication Date

May 3, 2024

Volume

13

Issue

13

Start / End Page

2323 / 2334

Related Subject Headings

  • 5108 Quantum physics
  • 5102 Atomic, molecular and optical physics
  • 4018 Nanotechnology
  • 1007 Nanotechnology
  • 0906 Electrical and Electronic Engineering
  • 0205 Optical Physics