AI Seminar: Machine Learning meets Particle Physics

Speaker

Troels Petersen, Associate Professor at Niels Bohr Institute, University of Copenhagen.

Abstract

Particle physics is renown for its complexity and enormous amounts of data - CERN’s LHC experiments has so far produced 250+ PB, after throwing away a factor 100000 to get it down in size! The reconstruction and identification of physics objects are well suited for Machine Learning approaches, and I will present our attempt at recognising electrons and photons in the proton collisions, and the estimation of their energy, both of which can significantly improve both the ATLAS Higgs-particle physics program, but also a whole row of other analysis.

See slides from the seminar here.