Les commentaires sont clos.

Random matrix advances in large dimensional statistics and machine learning

Date : 14-11-2017
Lieu : Université Paris V, Salle du conseil -- Espace Turing, 45 rue des Saint-Pères, Paris.

Thèmes scientifiques :
  • A - Méthodes et modèles en traitement de signal

Nous vous rappelons que, afin de garantir l'accès de tous les inscrits aux salles de réunion, l'inscription aux réunions est gratuite mais obligatoire.

S'inscrire à la réunion.


2 personnes membres du GdR ISIS, et 3 personnes non membres du GdR, sont inscrits à cette réunion.

Capacité de la salle : 50 personnes.


More up-to-date information and (free but mandatory) subscription at: https://www.ceremade.dauphine.fr/dokuwiki/mega:bigdataday2017

The recent application-driven surge of interest for the automated treatment of large dimensional datasets sets new challenges to (conventionally small dimensional) statistics and machine learning algorithms. The current ?large-p large-n? data paradigm alters the behavior of these algorithms, sometimes dramatically so, but at the same time enables the analysis of yet unseizable methods. This is the case notably of techniques involving non-linear statistics, such as kernel methods, random projections, but also community detection on large dimensional graphs, etc.

The objective of this GdR MEGA--GdR ISIS day is to share expertise and directions of future explorations in the field of applied random matrix theory and large dimension statistics alike, ranging from profoundly theoretical advances to more down-to-earth applications, notably to machine learning. A further target is to familiarize the signal and data processing community with these non-standard methods and to engage discussions on shared interests.

Confirmed speakers:

  • Alexandre D?Aspremont (Regularized non-linear acceleration)
  • Marc Lelarge (Fundamentals of low rank matrix estimation)
  • Abla Kammoun (A random matrix approach to discriminant analysis)
  • Edgar Dobriban (Factor selection by permutation in PCA and factor analysis)
  • Iain Johnstone (Random matrices for variance components models)
  • Romain Couillet (Random matrices in machine learning)
Organizers: Florent Benaych-Georges, Romain Couillet, Jamal Najim