Next-Value Prediction of Beam Waists From Mixed Pitch Grating Using Simplified Transformer Model

Next-Value Prediction of Beam Waists From Mixed Pitch Grating Using Simplified Transformer Model

Blog Article

In this study, a simplified transformer pinata cookie cutter model is used to predict the beam waist of 1,092 nm light coupled out from SiN-based mixed pitch gratings at various heights.The beam waists data at various heights above the grating is first compiled.Then, we used a sequence of the current beam waist values, z-positions, and the computed mathematical indicators (features) to predict the next beam waist value (labels).Optimized transformer model yields average percentage error (APE) of 6.6% between the predicted and actual beam waists, which corresponds to 93.

4% prediction accuracy.This study provides a angilina white pioneering approach to using natural language processing model to perform predictive modelling on photonics data, and possible extrapolation of photonics data using transformer model.