PNuT Seminar: Advances in developing deep neural networks for finding primary vertices in proton-proton collisions at the LHC

Abstract: In this presentation, I will delve into the exploration of deep neural networks (DNNs) for the identification and localization of primary vertices (PVs) in proton-proton collisions at the LHC. This study makes use of ATLAS simulated data at sqrt(s) = 13TeV, focusing on a pile-up range of 40-60, representative of LHC Run-3 conditions. The methodology employed follows a hybrid approach, starting with the estimation of kernel densities from the ensemble of charged track parameters. These kernel densities, coupled with truth vertices information, serve as input to a deep neural network that predicts the actual positions of primary vertices. The efficacy of this algorithm has been validated against the standard vertex finder algorithm utilized in ATLAS. I will present results showing a comparison of this algorithm with the ATLAS vertex finder and conclude with a brief overview of the ongoing progress in this project.