1-D band structure prediction in photonic fishbone structure using artificial neural network

We aim to train an artificial neural network to predict the band structure of 1-D fishbone photonic nanobeam and find the slow light mode in the structure. Using the geometric parameters, reduced k-vectors in reduced Brillouin zone, and the eigen-order as the input training dataset, the eigen-frequencies of the photonic modes can be predicted. The group velocity and the eigenfrequency of the slow light mode can be predicted by only input the desired conditions to the trained neural network.