Designing rectangular surface Bragg gratings using machine learning models

In this paper, we demonstrate a possibility to predict the characteristics of semiconductor-based Bragg gratings using machine learning methods. We perform 2D simulations of the Bragg gratings and calculate reflectance to create a database. With obtained data, we train ML models to predict the shape of the upper part of the main peak of reflectance. We compare the performance of the widely used neural network with various different models on our data and demonstrate the high accuracy of the optimized XGBoost method.