Machine Learning Application for Modeling and Design Optimization of High Frequency Structures

A brief review of the applications of machine learning to the electromagnetic modeling and design optimization of high‐frequency structures is presented. The structure of artificial neural networks (ANNs), their training, and testing phases are discussed. The applications of ANNs to the forward and inverse modeling of electromagnetic structures are presented. Machine learning is applied to accelerate electromagnetic modeling methods such as Method of Moments (MoM), Finite Difference Time‐Domain (FDTD), and variational methods. Finally, emerging applications of machine learning in unsupervised electromagnetic modeling and some future search directions are highlighted.