Würzburg 2018 – scientific programme

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

T: Fachverband Teilchenphysik

T 58: Top-Quarks: Eigenschaften II

T 58.3: Talk

Wednesday, March 21, 2018, 17:00–17:15, Z6 - SR 1.002

Probing the ttγ process at √s = 13 TeV with ATLAS using object and event based neural networks — Andreas Kirchhoff, Thomas Peiffer, Arnulf Quadt, Elizaveta Shabalina, •Joshua Wyatt Smith, Royer Edson Ticse Torres, and Knut Zoch — II. Physikalisches Institut, Georg-August-Universität Göttingen

Through the associated production of the tt+γ process we can measure the strength of the electromagnetic coupling of the top quark and the photon. Any deviation from the Standard Model (SM) prediction would be an indication of Beyond SM physics. Evidence of this process was seen at CDF with √s = 1.96 TeV, while observation occurred at the LHC at √s = 7 and √s =8 TeV, with increasing precision. In the 13 TeV analysis a requirement that photons be isolated is introduced. Thus, the most significant backgrounds in the single lepton channels come from electrons misidentified as prompt photons, as well as misidentified photons from hadrons or hadronic decays. In the dilepton channels the largest backgrounds are prompt photons from single top processes and Z→ l⁺ l⁻ γ decays. A neural network is trained on signal and the sum of backgrounds using object and event level information to maximise the separation. This response is used as the discriminating variable in the maximum likelihood fit to extract the ttγ cross-section. The event and object based neural network is presented, with initial fit results and a discussion on the largest systematic uncertainties.