The reconstruction of four-top-quark events in opposite-sign dilepton final state using Boosted Decision Trees and Neutrino Weighting

Abstract

A study of the reconstruction of opposite-sign dilepton four-top-quark events with ATLAS simulated samples using Boosted Decision Trees and neutrino weighting is presented. Boosted Decision Trees were used to determine correct assignments between jets and their truth origins, which is crucial for the full reconstruction of the four-top-quark events due to many jets in the decay products. The performance was analysed by determining the reconstruction efficiency that all the jets matched to the correct partons. An overall reconstruction efficiency of 29.1% was achieved. Neutrino weighting was applied to reconstruct the two leptonically decaying top quarks by determining the kinematics of the neutrinos. Its performance was analysed by determining the fraction of events that have solutions of the kinematics of the neutrinos, which has been achieved at 92.6%, and comparing the truth and reconstructed kinematics of leptonically decaying top quarks.

Date of Award	1 Aug 2021
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Evelina Gersabeck (Supervisor) & Yvonne Peters (Supervisor)