BIDS Machine Learning and Science Forum — Point cloud applications to collider physics

ML&Sci Forum

September 13, 2021
11:00am to 12:00pm
Virtual Participation

Register

BIDS Machine Learning and Science Forum
Date: Monday, September 13, 2021
Time: 11:00 AM - 12:00 PM Pacific Time
Location: Attend via Zoom 

Point cloud applications to collider physics

Speaker: Vinicius Mikuni, NERSC
Abstract: At the LHC Experiment, each proton collision creates thousands of particles. Extracting information from a high dimensional space such as the space of collisions requires algorithms that take advantage of particle symmetries while being capable of handling high dimensional inputs. Point cloud processing methods, often applied to robotics and self-driving cars, are able to handle such datasets and exploit the geometrical relationship between points. In this talk, I will present the application of this concept to different problems in collider physics, comparing the results with other well established algorithms.

The BIDS Machine Learning and Science Forum meets biweekly to discuss current applications across a wide variety of research domains in the physical sciences and beyond. Hosted by BIDS Affiliates Uroš Seljak (professor of Physics at UC Berkeley) and Ben Nachman (physicist at Lawrence Berkeley National Laboratory), these active sessions bring together domain scientists, statisticians, and computer scientists who are either developing state-of-the-art methods or are interested in applying these methods in their research.  To receive email notifications about upcoming meetings, or to request more information, please contact the organizers at berkeleymlforum@gmail.comAll interested members of the UC Berkeley and Berkeley Lab communities are welcome and encouraged to attend. 

Speaker(s)

Vinicius Mikuni

NERSC

Vinicius Mikuni is a new NESAP for Learning postdoc with NERSC working on machine learning applications to science. They recently finished their PhD in experimental particle physics in the University of Zurich, working with the CMS Experiment.