Sitzungen | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

FM: Fall Meeting

FM 82: Quantum & Information Science: Neural Networks, Machine Learning, and Artificial Intelligence III

FM 82.1: Invited Talk

Donnerstag, 26. September 2019, 14:00–14:30, 3044

Deep Learning Advances in Particle Physics — •Yannik Rath¹, Martin Erdmann¹, Benjamin Fischer¹, Erik Geiser¹, Jonas Glombitza¹, Dennis Noll¹, Thorben Quast^1,2, and Marcel Rieger¹ — ¹III. Physikalisches Institut A, RWTH Aachen University — ²EP-LCD, CERN

Machine learning methods have found widespread use in high-energy particle physics, their most common application being the identification of particles and the separation of signal and background processes in collision events. Deep learning in particular has seen many recent developments, for example the creation of dedicated neural network architectures incorporating physics knowledge (e.g. JINST 14 (2019) P06006). In addition, increasing attention has been directed towards unsupervised learning methods. Most notably, generative adversarial networks are extensively studied for their potential to speed up event simulations by several orders of magnitude (e.g. T. Comput Softw Big Sci 3 (2019) 4). Further unsupervised approaches based on reinforcement learning are also starting to be investigated. In this talk we present an overview of deep learning applications in high-energy particle physics focusing on most recent advancements.