Machine learning for knowledge acquisition and accelerated inverse-design for non-Hermitian systems

oleh: Waqas W. Ahmed, Mohamed Farhat, Kestutis Staliunas, Xiangliang Zhang, Ying Wu

Format: Article
Diterbitkan: Nature Portfolio 2023-01-01

Deskripsi

Machine learning has demonstrated effectiveness in optimizing complex physical structures. In this study, the authors employ a machine learning approach to inversely design non-Hermitian layered optical systems with gain and loss modulation, showing that the trained network can reveal the relation between asymmetric transmission and reflection spectra.