Tuesday, 19 October 2021 at 16:00 on Zoom

In this talk, I explain how various concepts and techniques in quantum field theory and collider physics can be naturally translated into a new geometric language. Using the energy mover's distance, which quantifies the minimal amount of "work" required to rearrange one event into another, we can define a distance between pairs of collider events. This distance can then be used to triangulate the "space" of collider events and rigorously define various geometric objects. Many well-known collider observables, jet algorithms, and pileup mitigation schemes have a simple geometric interpretation, as does the important concept of infrared and collinear safety.  This work helps illuminate the emerging connections between quantum field theory and machine learning.

Slides (PDF, 29 MB)