Leveraging long short-term memory (LSTM)-based neural networks for modeling structure-property relationships of metamaterials from electromagnetic responses

Prajith Pillai; Parama Pal; Rinu Chacko; Deepak Jain; Beena Rai

doi:10.1038/s41598-021-97999-6

Leveraging long short-term memory (LSTM)-based neural networks for modeling structure-property relationships of metamaterials from electromagnetic responses

Sci Rep. 2021 Sep 20;11(1):18629. doi: 10.1038/s41598-021-97999-6.

Authors

Prajith Pillai¹, Parama Pal², Rinu Chacko¹, Deepak Jain¹, Beena Rai¹

Affiliations

¹ TCS Research, Tata Consultancy Services, Mumbai, India.
² TCS Research, Tata Consultancy Services, Mumbai, India. parama.pal@tcs.com.

Abstract

We report a neural network model for predicting the electromagnetic response of mesoscale metamaterials as well as generate design parameters for a desired spectral behavior. Our approach entails treating spectral data as time-varying sequences and the inverse problem as a single-input multiple output model, thereby compelling the network architecture to learn the geometry of the metamaterial designs from the spectral data in lieu of abstract features.