Göttingen 2025 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
T: Fachverband Teilchenphysik
T 103: Methods in Particle Physics V (Event Reconstruction, PID)
T 103.3: Talk
Friday, April 4, 2025, 09:30–09:45, VG 4.101
Machine Learning Assisted Reconstruction of Hadron-Collider Events using Mini-Jets — •Josef Murnauer1, Stefan Kluth1, Daniel Britzger1, and Roman Kogler2 — 1Max-Planck-Institut für Physik, Garching — 2DESY, Hamburg
Reconstructing impactful physical observables from hadron collider data represents challenges due to combinatorial ambiguities and experimental effects. We propose a novel approach using mini-jets (R=0.1) as the sole reconstructed objects, employing a deep neural network for observable determination. This method condenses full event information into a manageable size, demonstrating superior efficiency and generality compared to classical algorithms for future LHC analyses.
Keywords: Machine Learning; Top Quark Physics; Reconstruction