Webrelaunch 2020

Numerical Analysis of Neural Networks (Summer Semester 2024)

The goal of the lecture is to provide a mathematical foundation of neural networks from the perspective of numerical analysis. Besides basic definitions and terminology, approximation properties are presented and connections to finite element methods are drawn. Further, numerical methods regarding the efficient training of neural networks are discussed and analyzed. The lecture also covers popular applications in the context of PDEs such as physics-informed neural networks.

Students who would like to attend the lecture are expected to have a solid background in numerical mathematics. Basic knowledge on functional analysis and finite element methods is also helpful, but not required.

Lecture: Tuesday 14:00-15:30 20.30 SR 2.067
Problem class: Wednesday 9:45-11:15 20.30 SR -1.012
Lecturer JProf. Dr. Roland Maier
Office hours: by appointment
Room 3.009 Kollegiengebäude Mathematik (20.30)
Email: roland.maier@kit.edu
Problem classes Dipl.-Math. Felix Krumbiegel
Office hours: by appointment
Room 3.005 Kollegiengebäude Mathematik (20.30)
Email: felix.krumbiegel@kit.edu