Collider experiments probe physics at the shortest distances by smashing protons together and measuring the debris produced by the collisions. However, it is challenging to infer the detailed short-distance physics from the distribution of debris measured by the detector. Over the past two years, the mathematical field of optimal transport has emerged as a promising tool for classifying collider events based on these distributions. In this talk, we will describe how optimal transport distances endow collider data with a geometric structure that can be used for ML-based event classification, and we will show how selecting optimal transport distances with good geometric properties can dramatically reduce computational effort.
Speakers: Katy Craig and Nathaniel Craig, UC Santa Barbara
Contact:Website: Click to Visit
Save this Event:iCalendar
Windows Live Calendar