A study of the four top production with deep neural network in pp collisions at √s=13 TeV in CMS
- 주제(키워드) Particle Physics , Deep Neural Network , Top Quark , Generative Adversarial Network (GAN)
- 발행기관 고려대학교 대학원
- 지도교수 최수용
- 발행년도 2021
- 학위수여년월 2021. 8
- 학위구분 박사
- 학과 대학원 물리학과
- 세부전공 입자물리학 전공
- 원문페이지 151 p
- UCI I804:11009-000000251865
- DOI 10.23186/korea.000000251865.11009.0001274
- 본문언어 영어
초록/요약
We study the standard model (SM) four top quark (tttt) productions in proton-proton collisions with root s = 13 TeV at the LHC using deep neural networks (DNNs). We use the data set corresponding to an integrated luminosity of 35.9 fb^-1 taken by the CMS detector in 2016. We select events which contain a pair of same sign di-lepton or at least three leptons (l = e or 𝜇). We adopt the multilayer perceptron (MLP) model to distinguish tttt events and the SM background events. The number of events observed after all selection requirements is consistent with the expectations from the background predictions. We set an upper limit on the cross section for tttt production in the SM of 19.6 fb at 95\% confidence level (CL, the signal strength of 2.1), with an expected limit of 19.2^+11.4_-6.8 fb. We use data-driven events generated by Wasserstein generative adversarial network (WGAN) for the expected background distributions, we set the upper limit on the cross section of 19.5 fb at 95\% CL (the signal strength of 2.1) with the expected limit of 18.5^+10.8_-6.4 fb.
more목차
1 Introduction 1
2 Theory 3
2.1 The Standard Model of Particle Physics ............. 3
2.1.1 Fermions and bosons .................... 3
2.1.2 Quantum electrodynamics ................. 5
2.1.3 Quantum chromodynamics................. 6
2.1.4 Weak interactions...................... 7
2.1.5 The standard model lagrangian .............. 10
2.2 Top Quark Physics ......................... 11
2.2.1 Four top quark production................. 11
3 The Apparatus 14
3.1 LargeHadronCollider ....................... 14
3.2 Compact Muon Solenoid detector ................. 16
3.2.1 Coordinate system ..................... 17
3.2.2 Solenoid magnet ...................... 18
3.2.3 Inner tracking system.................... 19
3.2.4 Electromagnetic calorimeter ................ 20
3.2.5 Hadronic calorimeter .................... 22
3.2.6 Muon system ........................ 24
3.2.7 Trigger and Data acquisition system ........... 26
3.3 The new Gamma Irradiation Facility ............... 27
3.3.1 The KODEL iRPC ..................... 27
3.3.2 An algorithm for clustering and tracking ......... 31
4 Event Reconstruction and Object Selection 36
4.1 Track reconstruction ........................ 36
4.2 Primary vertices........................... 37
4.3 Particle flow............................. 37
4.4 Lepton multi-isolation ....................... 39
4.5 Muons................................ 42
4.6 Electrons............................... 43
4.7 Jets ................................. 45
4.8 b-tagging .............................. 46
4.9 Missing transverse energy ..................... 47
5 Analysis Techniques 48
5.1 The deep neural networks ..................... 48
5.1.1 The multi layer perceptrons ................ 48
5.1.2 The Generative Adversarial Network ........... 49
5.2 Systematic uncertainties ...................... 51
5.3 Higgs Combined Tool........................ 52
6 Analysis 56
6.1 Data and Simulation Samples ................... 56
6.1.1 Collision data........................ 57
6.1.2 Monte Carlo simulations .................. 58
6.2 Triggers ............................... 65
6.3 Event Selection ........................... 66
6.3.1 The Same-Sign Di-Lepton Analysis ............ 66
6.3.2 The Tri-or-More Lepton Analysis ............. 67
6.4 Scale factors............................. 67
6.5 Event yields............................. 68
6.5.1 The Same-Sign Di-Lepton Analysis ............ 68
6.5.2 The Tri-or-More Lepton Analysis ............. 70
6.6 Control Plots ............................ 72
6.7 The MLP discriminator analysis.................. 73
6.8 Systematic uncertainties ...................... 76
6.9 Limit settings ............................ 78
7 WGAN analysis 81
7.1 Event generation .......................... 81
7.2 The MLP discriminator analysis.................. 83
7.3 Limit setting ............................ 83
8 Conclusion 85
A Control Plots 86
A.1 The Same-Sign Di-Lepton Analysis ................ 86
A.2 The Tri-or-More Lepton Analysis ................. 95
B Systematic Plots 101
B.1 The Same-Sign Di-Lepton Analysis ................102
B.2 The Tri-or-More Lepton Analysis .................108
C The WGAN generated Plots 114
C.1 The Same-Sign Di-Lepton Analysis ................115
C.2 The Tri-or-More Lepton Analysis .................123