Back to search

Aristotle University Of Thessaloniki (AUTH)

Application of statistical methods with emphasis on Machine Learning for the analysis of experimental data from the ATLAS experiment at the LHC at CERN

Abstract

dc:description

The present dissertation studies mainly interactions of the W and Z bosons using data from the ATLAS detector collected during the operation of CERN's LHC accelerator in the period 2015-2018, also "Run 2" operating period. This data corresponds to an integrated luminosity of 139 fb-1. This study is carried out on the one hand in the general context of the WZ diboson production, and on the other hand in the context of the Vector Boson Scattering (VBS) interactions, characterized by the simultaneous presence of two hadronic jets in the final state, along with the two vector bosons. In the beginning, an introduction to the ATLAS experiment and to the experimental methodology used for data collection and analyses is presented. Then, work performed as part of the authorship qualification task for the reception of the ATLAS author status is presented. This work entails firstly the author's contribution to the calculation of scale factors to be used in simulated muons in the forward part of the ATLAS detector, in the rapidity range of 2.5<|η|<2.7 . These factors are used to correct available simulated data to be used in experimental analyses. In parallel, responsibilities in the quality assurance-quality control were taken up during the MicroMegas construction procedure, that was realised at the Particle Detector Construction Laboratory, in the Physics Department at A.U.Th. The MicroMegas detectors have already been installed in the experiment as part of the Muon Spectrometer and specifically the New Small Wheel (NSW). A description of the physics motivation and the experimental procedure followed for the analysis is presented in the next chapters. First, the calculation of Reducible Background in the Diboson WZ production is studied. This background consists of the ttbar, Z+jets and Zγ processes, and it can be mistaken as signal due to one or more of the selected leptons is a mis-identified jet or lepton coming from photon conversion, called 'fake' leptons in this context. To avoid directly using inaccurate Monte Carlo (MC) simulation to estimate this background, a data-driven computation is performed using a method that involves the inversion of a generalized matrix containing probabilities of selecting a 'fake' (Fake Factors), computed in separate Control Regions using MC simulation, and the normalization factor is computed. Next, the expected discovery significance of the purely-electroweak WZ process is measured in the WZjj VBS phase space. In addition, the normalization factor for WZ-EW simulation is calculated. The Profile Likelihood method was used to simultaneously fit signal and background control regions, utilizing the full Run 2 available simulated events. Finally, results on the parameter estimation from a search of signs of New Physics using an Effective Field Theory re-interpretation of the WZ electroweak process in the WZjj VBS phase space are presented. The search relates to deviations from the Standard Model of Quartic Gauge Boson (QGCs) interactions, with relevant effective field operators of dimension 8. The confidence limits of allowed parameter values for the model's Lagrangian operator coefficients are calculated using Run 2 simulated and observed data.

Degree

thesis:*
Grantor dc:publisher
Aristotle University Of Thessaloniki (AUTH)
Year dc:date
2022

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Karkanias, Ioannis
  • Καρκανιάς, Ιωάννης

Subjects

dc:subject × 12

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/51705

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Karkanias, Ioannis; Καρκανιάς, Ιωάννης. Application of statistical methods with emphasis on Machine Learning for the analysis of experimental data from the ATLAS experiment at the LHC at CERN. Aristotle University Of Thessaloniki (AUTH), 2022. http://hdl.handle.net/10442/hedi/51705