Calorimetry with deep learning: particle simulation and reconstruction for collider physics
- PDF / 7,542,695 Bytes
- 31 Pages / 595.276 x 790.866 pts Page_size
- 15 Downloads / 178 Views
Regular Article - Experimental Physics
Calorimetry with deep learning: particle simulation and reconstruction for collider physics Dawit Belayneh1, Federico Carminati2 , Amir Farbin3 , Benjamin Hooberman4 , Gulrukh Khattak2,5 , Miaoyuan Liu6 , Junze Liu4 , Dominick Olivito7 , Vitória Barin Pacela8 , Maurizio Pierini2 , Alexander Schwing4 , Maria Spiropulu9 , Sofia Vallecorsa2 , Jean-Roch Vlimant9 , Wei Wei4 , Matt Zhang4,a 1
University of Chicago, Chicago, IL, USA European Organization for Nuclear Research (CERN), Geneva, Switzerland 3 University of Texas Arlington, Arlington, TX, USA 4 University of Illinois at Urbana-Champaign, Champaign, IL, USA 5 UET Peshawar, Peshawar, Pakistan 6 Fermi National Accelerator Laboratory, Batavia, IL, USA 7 University of California, San Diego, CA, USA 8 University of Helsinki, Helsinki, Finland 9 California Institute of Technology, Pasadena, CA, USA
2
Received: 8 January 2020 / Accepted: 16 July 2020 © The Author(s) 2020
Abstract Using detailed simulations of calorimeter showers as training data, we investigate the use of deep learning algorithms for the simulation and reconstruction of single isolated particles produced in high-energy physics collisions. We train neural networks on single-particle shower data at the calorimeter-cell level, and show significant improvements for simulation and reconstruction when using these networks compared to methods which rely on currently-used state-of-the-art algorithms. We define two models: an endto-end reconstruction network which performs simultaneous particle identification and energy regression of particles when given calorimeter shower data, and a generative network which can provide reasonable modeling of calorimeter showers for different particle types at specified angles and energies. We investigate the optimization of our models with hyperparameter scans. Furthermore, we demonstrate the applicability of the reconstruction model to shower inputs from other detector geometries, specifically ATLAS-like and CMS-like geometries. These networks can serve as fast and computationally light methods for particle shower simulation and reconstruction for current and future experiments at particle colliders.
a e-mail:
[email protected] (corresponding author)
0123456789().: V,-vol
1 Overview In high energy physics (HEP) experiments, detectors act as imaging devices, allowing physicists to take snapshots of final state particles from collision “events”. Calorimeters are key components of such detectors. When a high-energy primary particle travels through dense calorimeter material, it deposits its energy and produces a shower of secondary particles. Detector “cells” within the calorimeter then capture these energy depositions, forming a set of voxelized images which are characteristic of the type and energy of the primary particle. The starting point of any physics analysis is the identification of the types of particles produced in each collision and the measurement of the momentum carried by each of these particles. These tasks hav
Data Loading...
