BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Renormalisation group and machine learning: the Wavelet-Conditiona
 l RG  - Giulio Biroli (ENS Paris)
DTSTART:20230516T120000Z
DTEND:20230516T130000Z
UID:TALK199942@talks.cam.ac.uk
CONTACT:Sarah Loos
DESCRIPTION:Reconstructing\, or generating\, high dimensional distribution
 s starting from data is a central problem in machine learning and data sci
 ences.\nI will present a method —The Wavelet Conditional Renormalization
  Group —that combines ideas from physics (renormalization group theory) 
 and computer science (wavelets\, stable representations of operators). The
  Wavelet Conditional Renormalization Group allows to reconstruct in a very
  efficient way classes of high dimensional probability distributions hiera
 rchically from large to small spatial scales\, and to perform RG directly 
 from data.  It allows to bridge the gap between approaches based on physic
 al intuition and modern machine learning algorithms. I will present the me
 thod and then show its applications to data from statistical physics and c
 osmology. I shall also discuss the interesting insights that our method of
 fers on the interplay between structures of data and architectures of deep
  neural networks.\n
LOCATION:Center for Mathematical Sciences\, Lecture room MR4
END:VEVENT
END:VCALENDAR