The Scientific ML team of the Machine Learning Journal Club has recently worked with the Julia library #NeuralPDE.jl, a useful tool for the automated resolution of differential equations using Physics-Informed Neural Networks #PINNs.
In particular, they have produced a series of examples of PDEs from relevant physics problems in order to put together a detailed benchmark of the performance of the library at its current state of development.
We are honoured to have contributed to the first comprehensive paper on NeuralPDE.jl together with Chris Rackauckas and other authors. You can find it on arXiv: https://arxiv.org/ftp/arxiv/papers/2107/2107.09443.pdf
We hope it will be a useful resource for people interested in using NeuralPDE.jl for their projects.
#ai #machinelearning #pinns