BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Sampling using Diffusion Processes - Dr Nicolas Macris\, EPFL
DTSTART:20250212T140000Z
DTEND:20250212T150000Z
UID:TALK225808@talks.cam.ac.uk
CONTACT:Prof. Ramji Venkataramanan
DESCRIPTION:I will discuss a class of diffusion-based algorithms to draw s
 amples from high-dimensional probability distributions given their unnorma
 lized densities. Ideally\, the method can transport samples from a Gaussia
 n distribution to a specified target distribution in\nfinite time. The sto
 chastic interpolants framework used to\nderive a diffusion process\, and a
 lso involves solving certain Hamilton-Jacobi-Bellman PDEs. These are solve
 d using the theory of forward-backward stochastic differential equations (
 FBSDE) together with machine learning-based methods. Numerical experiments
  illustrating that the algorithm will also be discussed.\nThis is joint wo
 rk with Anand Jerry George.
LOCATION:MR5\, CMS Pavilion A
END:VEVENT
END:VCALENDAR
