Penn researchers have developed a smarter AI method for solving notoriously difficult inverse equations, which help ...
Abstract: In this letter, we explore the feasibility of using machine learning (ML) to address time-domain electromagnetic forward and inverse problems. A novel ML-associated Hermite (AH) ...
Abstract: Based on unsupervised physics-informed neural network (PINN) framework, a two-dimensional inverse-design method for antenna superstrate is proposed, which can simultaneously realize the ...