Johann Bernoulli lezing 34

Jürgen Jost
MPI for Mathematics in the Sciences, Leipzig

Title: Curvature: How a simple geometric concept became fundamental for theoretical physics and inspires machine learning

Abstract:

Curvature was originally conceived as a measure for curves in the plane and surfaces in spaces. In the hands of Gauss and Riemann, it became an intrinsic invariant of a Riemannian metric, with profound applications in Einstein's theory of general relativity, in quantum field theory, and of course, in many branches of mathematics. It was then found that curvature can be characterized by metric relations in triangles, and therefore defined for spaces that are more general than Riemannian manifolds. In our recent research, links with concepts from topological data analysis were discovered, leading to a new conceptualization of curvature with potential applications in machine learning.

Laudatio: tba

The lecture can also be followed on youtube (link will follow soon)