Erlangen 2026 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

T: Fachverband Teilchenphysik

T 46: Top Physics II

T 46.3: Talk

Wednesday, March 18, 2026, 16:45–17:00, KH 00.011

Using Machine Learning Techniques for a Search for Single Top Quark Production — •Niklas Düser and Andrea Knue — TU Dortmund

The search for single-top-quark production in the s-channel is experimentally challenging due to its small cross-section and similarity to dominant background processes. This analysis investigates two machine-learning approaches to enhance signal discrimination: a deep neural network (DNN) using high-level kinematic variables, and a graph neural network (GNN) encoding event topology through particle-object correlations. Both models were trained and evaluated on simulated proton-proton collisions at a center-of-mass energy of 13 TeV, corresponding to the Run 2 dataset at ATLAS with an integrated luminosity of 140 fb-1. The performance of the DNN and GNN is compared with a focus on signal and background modelling uncertainties.

Keywords: s-channel; single top; cross section; top