The goal of this project is to provide an introduction to physics-informed neural networks (PINNs). For phenomena that can be mathematically described by certain differential equations, PINNs establish a predictive modeling approach that is both physics-guided and data-driven.
Two notebooks provide a concise introduction and a practical demonstration on the basis of a one-dimensional time-dependent heat transfer problem.
Everything here is work in progress at this point.
