H.B. Keller Colloquium

Gradient flows for sampling and their deterministic interacting particle approximations

Dejan Slepčev, Professor, Department of Mathematical Science, Carnegie Mellon University,

Speaker's Bio:
Dejan Slepčev is a Professor at the Department of Mathematical Sciences of Carnegie Mellon University. His research interests include applied analysis, partial differential equations, optimal transportation, calculus of variations and applications to problems of data science and physical sciences.

Motivated by the task of sampling measures in high dimensions we will discuss a several gradient flows in the spaces of measures, including the Wasserstein gradient flows of Maximum Mean Discrepancy and relative entropy, the Stein Variational Gradient Descent and a new Radon-Wasserstein gradient flows. For all the flows we will consider their deterministic interacting-particle approximations. The talk will highlight some of the properties of the flows and indicate their differences. In particular we will discuss how well can the interacting particles approximate the target measures. The talk is based on joint works with Elias Hess-Childs, Anna Korba, Sangmin Park, Lihan Wang, and Lantian Xu.

Event Sponsors:

Computing and Mathematical Sciences (CMS) - More events from this Sponsor

For more information, please contact Sumaia Abedin by phone at 6263956704 or by email at sabedin@caltech.edu or visit https://www.cms.caltech.edu/news-events/keller-colloquium.

Add this event to your calendar.