Die DPG-Frühjahrstagung in Bonn musste abgesagt werden! Lesen Sie mehr ...
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
T: Fachverband Teilchenphysik
T 79: Experimental methods IV
T 79.1: Vortrag
Donnerstag, 2. April 2020, 16:30–16:45, H-ÜR 1
Fast simulation of Electromagnetic Calorimeter using Variational Autoencoder — •Jubna Irakkathil Jabbar2, Florian Bernlochner1, Pablo Goldenzweig2, and Jochen Gemmler2 — 1University of Bonn, Germany — 2Karlsruhe Institute of Technology, Germany
The simulation of particle showers in electromagnetic calorimeters with high precision is a computationally expensive and time consuming process. Fast simulation of particle showers using generative models have been suggested to significantly save computational resources. In this study, electron showers simulated using the Geant4 simulation toolkit are used to train a variational autoencoder model. The model consists of an encoder, decoder and a latent noise vector. Once the model is trained, the decoder is used to generate particle shower simulations providing noise vectors as input. The generated particle showers are cross-checked with the Geant4 showers using various observables.