Lecture SS 26 Advanced Topics in Scientific Computing

Mathematics of Machine Learning and Shape Space Theory

Lecturer
Prof. Martin Rumpf

Basis eCampus

Topics covered by the lecture course:

  • neural networks for the approximation of functions and geometries,
  • distances in the space of images, curves and surfaces,
  • neural networks for the solution of PDEs in a geometric context,
  • autoencoders and their regularization,
  • geometric calculus on latent spaces.

Prerequisites: basic knowledge of functional analysis and the finite element method.