Wuppertal 2015 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
T: Fachverband Teilchenphysik
T 20: Top: Boosted top, Resonanzen
T 20.1: Gruppenbericht
Montag, 9. März 2015, 16:45–17:05, L.10.31 (HS 10)
HEPTopTagger Performance with 2012 Data in ATLAS — Christoph Anders, Maddalena Giulini, Sebastian Schätzel, André Schöning, and •David Sosa — Physikalisches Institut, Universität Heidelberg
The HEPTopTagger algorithm (Plehn et al. JHEP 1010 (2010) 078) reconstructs hadronically decaying top quarks with transverse momentum pT > 200 GeV. The algorithm takes Cambridge/Aachen R=1.5 fat jets as input and removes contributions from underlying event and pile-up processes. It then tests combinations of subjets inside the fat jet for compatibility with the top quark and W boson masses.
The HEPTopTagger performance was evaluated with the full 2012 ATLAS data set using the lepton+jets channel (e and µ). Data/MC comparisons were carried out for distributions of kinematic variables of fat jets and subjets before and after applying the HEPTopTagger. All relevant systematic uncertainties have been evaluated, the most important being the subjet energy scale, the b-tagging and the tt normalisation uncertainties.
The signal efficiency was derived for background-subtracted data and in simulation as a function of fat jet η and pT. MC-to-Data efficiency scale factors were obtained including all systematic uncertainties and are ready to be applied in HEPTopTagger analyses. In addition, a dijet data sample was used to determine the mistag rate. The HEPTopTagger performance was compared to that of other top taggers using simulated tt and QCD dijet events.