• Lecturer: Guido Montúfar
• Date: Thursday 11:15 - 12:45
• Room: MPI MiS, A3 02
• First lecture: April 14th, 2016
• Language: English
• Target audience: MSc students, PhD students, Postdocs
• Content (Keywords): artificial neural networks, supervised learning, pattern classification, VC-dimension, theory of deep learning
• Prerequisites: basic linear algebra and analysis
• Literature: M. Anthony and P. Bartlett, Neural Network Learning: Theoretical Foundations, Part one.

Abstract

This lecture gives an introduction to neural network learning from a theoretical standpoint. The main focus is on supervised learning problems, covering topics on learnability, generalization, and complexity. The lecture will also touch on current developments on the theory of deep learning.