AI/ML+Physics Part 3: Designing an Architecture - presented by Prof. Steve Brunton

AI/ML+Physics Part 3: Designing an Architecture

Prof. Steve Brunton

Prof. Steve Brunton
Slide at 27:49
a(x) Fourier layer 1 Fourier layer 2 Fourier layer T u(x) Fourier layer v(x) Operator learning can be taken as an image-to-image problem. The Fourier layer can be viewed as a substitute for the convolution layer.
1
References
  • 1.
    Zongyi Li et al. (2020) Fourier Neural Operator for Parametric Partial Differential Equations.
Share slide
Summary (AI generated)

The Fourier neural Operator is based on the idea that real-world physics is multiscale and efficiently represented in the Fourier domain. By incorporating Fourier layers into the neural Operator, it implicitly assumes a multiscale nature of physics. Graph neural networks are another example of this concept.