Solving Differential Equations with Neural Networks

A hands-on introduction to physics-informed neural networks with PyTorch

Over the last decades, artificial neural networks have been used to solve problems in varied applied domains such as computer vision, natural language processing and many more. Recently, another very promising application has emerged in the scientific machine learning (ML) community: The solution of partial differential equations (PDEs) using artificial neural networks, using an approach normally referred to as physics-informed neural networks (PINNs). Today, PINNs …

differential-equations machine learning networks neural networks physics pytorch