BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Are generative models the new sparsity?
DTSTART:20190725T150000
DTEND:20190725T160000
DTSTAMP:20260507T185024Z
UID:afd335724151952eb690264b21363133bd00ea14ad89c39020703d8d
CATEGORIES:Conferences - Seminars
DESCRIPTION:Lenka Zdeborová\nAbstract:\nSparse principle component analys
 is is being used in a range of applications and attracted interest of theo
 reticians for its algorithmically challenging properties. Sparsity is a wi
 dely explored way to reduce dimensionality. Another such way\, that is rec
 ently widely studied\, is learning generative models from data. In this ta
 lk I will discuss what happens when sparsity is replaced by generative mod
 els. I will present a detailed study of the spiked matrix model with the s
 pike coming from generative model. I will show that the computational gap 
 well-known in sparse principle component analysis does not exist in this 
 case. I will discuss the behaviour of message passing algorithms and a con
 struction and analysis of optimality-achieving spectral algorithms. Talk i
 s based on arxiv:1905.12385.\n 
LOCATION:BC 420 https://plan.epfl.ch/?room==BC%20420
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR