Erlangen 2026 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

T: Fachverband Teilchenphysik

T 32: Higgs Physics IV

T 32.4: Vortrag

Dienstag, 17. März 2026, 17:00–17:15, KH 01.019

ML-driven Di-Tau Mass and Momentum Reconstruction at the ATLAS Detector — •Jonathan Pampel, Tatjana Lenz, and Jochen Dingfelder — University of Bonn

A common process in particle physics analyses is the decay of a neutral particle into two tau leptons. Each tau lepton decay contains at least one neutrino, making the reconstruction of the 4-momentum – and in particular, the mass – of the neutral particle an underdetermined problem.

In this work, a neural network has been trained on ATLAS γ^* Monte Carlo samples at √s=13.6 TeV to estimate the di-tau system’s 4-momentum, using the reconstructed visible tau 4-momenta and the missing transverse momentum as input features.

This talk presents the development of the neural network and its performance in comparison to the Missing Mass Calculator (MMC), the previously established tool for di-tau 4-momentum reconstruction.

Keywords: Tau Lepton; Missing Mass Calculator; MMC; Machine Learning; Ditau