Karlsruhe 2024 – scientific programme

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

T: Fachverband Teilchenphysik

T 84: Methods in particle physics 5 (tagging)

T 84.4: Talk

Thursday, March 7, 2024, 16:45–17:00, Geb. 20.30: 2.066

Calibration of boosted vector boson tagging algorithms in ATLAS using diboson events — •Simone Ruscelli and Chris M. Delitzsch — TU Dortmund

The unprecedented centre-of-mass energy of the pp collisions at the Large Hadron Collider enables vector bosons production with a transverse momentum much larger than their rest mass, resulting in the collimation (or boost) of their decay products. The hadronic decay products of the vector bosons are captured within a large-radius (R = 1.0) jet. To distinguish such a jet from a quark/gluon-initiated jet, boosted object identification (tagging) algorithms are developed that exploit the radiation pattern within the large-radius jet, also known as jet substructure. The state-of-the-art tagging algorithms are based on machine learning techniques and significantly outperform simple cut-based taggers using a few jet substructure variables. Before such tagging algorithms can be used in analyses, they must be calibrated to match the efficiency obtained in Monte Carlo Simulation to that in data. To calibrate boosted W/Z boson tagging algorithms, tt events and V+jets events are used. While V+jets events cover the p_T > 500 GeV range, W bosons from top-quark decays are only available up to p_T ≈ 350 GeV. In these feasibility studies, diboson events (WZ and ZZ) in which the Z-boson decays into a pair of charged leptons are explored to fill the gap in p_T between the tt and V+jets selections, but also to provide a cross-check to the current analyses in regions of overlap. The studies use √s = 13 TeV pp collisions recorded by the ATLAS detector at the Large Hadron Collider between 2015 and 2018.

Keywords: Boosted object tagging; Scale factors; Diboson events; Large-radius jets