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

T: Fachverband Teilchenphysik

T 24: Higgs Physics III (boson final states)

T 24.4: Talk

Tuesday, April 1, 2025, 17:00–17:15, ZHG104

Optimization of machine learning-based measurements of Higgs production processes in the H → 4ℓ decay channel with ATLAS Run 3 data — •Luca Spitzauer, Sandra Kortner, Hubert Kroha, Alice Reed, Elena Cuppini, and Tae Hyoun Park — Max-Planck-Institut für Physik

Cross-section measurements for various Higgs boson production and decay processes are crucial for exploring Higgs boson properties and have high sensitivity to potential physics beyond the Standard Model. The decay of a Higgs boson into a pair of Z bosons, each subsequently decaying into two leptons (H → ZZ^* → 4ℓ), is particularly important for these measurements due to its exceptionally clear signal.

Within the framework of Simplified Template Cross Sections (STXS), exclusive regions of phase space are defined for each Higgs boson production mode. Optimized classification of reconstructed events according to the STXS production regions is essential to enhance signal sensitivity and reduce uncertainties. The previous round of STXS measurements in the H → 4ℓ channel using the Run 2 ATLAS dataset employed a Neural Network classification approach. With the new Run 3 dataset at a center-of-mass energy of 13.6 TeV, we are exploring potential optimizations of this classification using a new Deep Set machine-learning approach.

Keywords: ATLAS; Higgs; Machine Learning; Cross-section; Lepton