2021/06/02 by Shah Rukh Qasim, K. Long, Qasim, Shah Rukh +7
Physics and Astronomy · #FOS: Physical sciences #Instrumentation and Detectors (physics.ins-det) #Particle Detector Development and Performance #Particle physics theoretical and experimental studies #Radiation Detection and Scintillator Technologies
paper · pdf · doi:10.48550/arxiv.2106.01832
openalex publication_date 2021/06/02 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
The high-luminosity upgrade of the LHC will come with unprecedented physics\nand computing challenges. One of these challenges is the accurate\nreconstruction of particles in events with up to 200 simultaneous proton-proton\ninteractions. The planned CMS High Granularity Calorimeter offers fine spatial\nresolution for this purpose, with more than 6 million channels, but also poses\nunique challenges to reconstruction algorithms aiming to reconstruct individual\nparticle showers. In this contribution, we propose an end-to-end\nmachine-learning method that performs clustering, classification, and energy\nand position regression in one step while staying within memory and\ncomputational constraints. We employ GravNet, a graph neural network, and an\nobject condensation loss function to achieve this task. Additionally, we\npropose a method to relate truth showers to reconstructed showers by maximising\nthe energy weighted intersection over union using maximal weight matching. Our\nresults show the efficiency of our method and highlight a promising research\ndirection to be investigated further.\n