BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Geometric Optimization in Scientific Machine Learning
DTSTART:20260225T133000
DTEND:20260225T150000
DTSTAMP:20260415T020559Z
UID:b1944f1efc1e28a6a709969d974db32942d5735d4a1d531b213cd867
CATEGORIES:Conferences - Seminars
DESCRIPTION:Marius Zeinhofer\nWe discusses an “optimize-then-project” 
 approach for applications in scientific machine learning. The key idea is 
 to design algorithms at the infinite-dimensional level and subsequently di
 scretize them in the tangent space of the neural network ansatz. We illust
 rate this approach in the context of the variational Monte Carlo method fo
 r quantum many-body problems\, where neural quantum states have recently e
 merged as powerful representations of high-dimensional wavefunctions. In t
 his setting\, we recover the celebrated stochastic reconfiguration algorit
 hm\, interpreting it as a projected Riemannian L2 gradient descent method.
  We further explore extensions to Riemannian Newton methods\, and conclude
  with considerations related to the scalability of these schemes.\n 
LOCATION:CM 1 517 https://plan.epfl.ch/?room==CM%201%20517
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR