Course Information

The lecture will discuss the Statistical Mechanics approach to understanding the general principles which underly the success of deep neural networks, including the analysis of Gibbs and online learning for teacher-student setups. The discussion will also include Hopfield networks, a foundational concept recently honored by the Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton.

Class Times: Tuesday 13:15-14:45,
R114, ITP, Brüderstraße 16 (Lecture)
  Wednesday 17:15-18:45,
R114, ITP, Brüderstraße 16 (Lecture)
  Monday 9:15-10:45,
R114, ITP, Brüderstraße 16 (Tutorial)
Recommended Reading: Michael Nielsen, Neural Networks and Deep Learning ;
  A. Engel and C. Van den Broeck, "Statistical Mechanics of Learning" ;
  Michael Biehl, The Shallow and the Deep ;
Results : Exam grades

Problem Sets and Lecture Notes

You can ask questions concerning the problem sets and tutorials to Marcel Kühn (mkuehn_at_itp.uni-leipzig.de, R. 311a). If you plan to visit the office please make an appointment in advance.


2024