Dynamical Aspects of Machine Learning

Dynamical Aspects of Machine Learning

Friday, January 18th 11:00-12:00, 2019
TIMON Room at DeustoTech

Yacine Chitour
Laboratoire des signaux et systèmes (L2S), CNRS – CentraleSupelec – Universitè Paris-Sud, France.

Abstract:

This talk is a brief and elementary introduction to the dynamical systems issues associated with the gradient descent algorithms involved in deep learning. Some results related to linear deep learning will be presented. This is a joint work with Z. Liao and R. Couillet.

See the slides here.

The recording of the seminar can be found below: