1 Optimizing Selection and Sensitivity Results for VV->lvqq, 6.5 pb, 13 TeV Supervisor: Dr. Kalliopi Iordanidou 215 Columbia University REU Home Institution: High Point University
2 Summary Introduction CERN and the Large Hadron Collider The ATLAS Experiment Particles and Monte Carlo Samples VV Resonance and Selection Resonance and Decay Selection Selection Optimization Selection & Extrapolations Significance Plots & Trigger Studies Multivariate Analysis Topology Plots Conclusions
CERN 3 In 1954 twelve countries ratified the European Council for Nuclear Research, CERN Located near Geneva, Switzerland along Franco- Swiss border Today there are 21 member states Focused on particle physics research Home to the Large Hadron Collider (LHC)
4 Large Hadron Collider Largest and most powerful particle accelerator in the world 27 kilometer ring that holds superconducting magnets High energy particle beams travel in opposite directions Collisions at 4 main particle detectors, ALICE ATLAS, CMS, LHCb Run 1: March 21-February 213 Run 2: April 215-present
The ATLAS Experiment 5 A Toroidal LHC ApparatuS, large general purpose particle detector 7, tonnes, 46m x 25m x 25m Four major components Inner Detector: measures momentum of charged particles Calorimeter: measures energies carried by particles Muon Spectrometer: identifies and measures momenta of muons Magnet System: bends charged particles for momentum measurement
Particles 6 Leptons Elementary particles, do not undergo strong interactions Six flavors, three generations First Generation: Electron e - & Electron Neutrino v e Second Generation: Muon µ - & Muon Neutrino v µ Third Generation: Tau τ & Tau Neutrino v τ For all generations e -, µ -, τ : negative charge, elementary particle v e, v µ, v τ : no electric charge Anti-Particles: Positron e +, Antimuon µ +, Antitau τ + Same mass as respective particle, opposite charge
Particles cont. 7 Gauge Bosons Type of vector boson Three types Photons: carry electromagnetic interaction W & Z Bosons: carry weak interaction Gluons: carry strong interaction Missing Transverse Energy (MET) Used to deduce presence of non-detectable particles (ex: neutrino)
Particles cont. 8 Hadrons Composite particles made of quarks Held together by strong force Two categories: Baryons & Mesons Baryons (ex: protons and neutrons) Made of three quarks Mesons (ex: pion) One quark, one antiquark Jets High-energy quark transformed into spray of hadrons Most commonly used in ATLAS Regular Jets:.4 distance parameter Large-R Jets: 1. distance parameter Boosted Jets
Resonance and Decay 9 Resonance Peak at a particular mass when two particles cross sections interactions are examined as a function of energy Cause: creation of a particle whose mass-energy is energy of resonance Described by mass spectrum of dispersion type Maximum = resonance mass m Spectrum width = probability of resonance decay Γ Does not exceed mc 2
Resonance and Decay cont. 1 VV Resonance See Feynman diagram, studied in lvqq final state Example of how signal bump should show above the background, if signal exists
and Monte Carlo Samples 11 6.5 pb at 13 TeV, Heavy Vector Triplet Background W + Jets Z + Jets ttbar No multijet or standard model diboson samples, estimated to be very small from Run 1 Run 1 Multijets Extrapolation 29 recorded events with luminosity of 2.3 fb at 8 TeV -> average of 6 events ± 3 events
Monte Carlo Scale 12 -> HVT 2 TeV MadGraphPythia
Selection 13 Tight Electron P T > 25 GeV, η < 2.47, tight working point Medium Muon P T > 25 GeV, η < 2.5, medium working point Level 1 Trigger (L1 Trigger) Lowest level trigger Boosted case Close Quarks Reconstructions identified as Large-R Jets
14 Selection cont. Baseline Selection Tight electron, medium muon, L1 Trigger MET > 3 GeV Lepton Neutrino P T > 1 GeV At least on Large-R Jet with P T > 1 GeV
Selection Optimization 15 Additional selection to study sensitivity of signal selection against background Increase in MET cuo MET > 1 GeV Clean selection of quantum chromo dynamics (QCD) contamination Large-R Jets mass between 65 GeV Large-R Jet P T > 2 GeV Lepton Neutrino P T > 2 GeV 8 7 6 5 4 3 2 1 L = ~6.5 pb, s = 13 TeV! 5 1 15 2 25 3 [GeV] M lvj
MET Plots 16 16 14 12 1 8 6 4 L = ~6.5 pb!, s = 13 TeV 14 12 1 8 6 4 L = ~6.5 pb!, s = 13 TeV 2 2 2 4 6 8 1 12 14 MET lvj [GeV] 1 15 2 25 3 35 4 45 5 55 6 MET lvj [GeV] MET with before cuts, low mass excess is QCD MET after > 1 GeV cut
Extrapolations of and Background at various luminosities 17 Reflective of how yields will progress as luminosity increases Expeco get up to 1 fb by the end of Run 2 ~3 years
Extrapolations of and Background at various luminosities 18
Significance Plots 19 Ratio of signal sample with respeco background is 2 TeV resonance signal!3.6 "1 ATLAS Work In Progress.5 L = ~6.5 pb, # s = 13 TeV.4.3.2 Significance plot of Large-R Jet mass between 65 GeV.1 5 1 15 2 25 3 35 M [GeV]
Significance Plots cont. Significance plot of Large-R Jet D2.25.2.15.1.5!3 "1.35 ATLAS Work In Progress.3 L = ~6.5 pb, # s = 13 TeV.5 1 1.5 2 2.5 3 3.5 4 4.5 5 D2 Significance plot of P lv T/M lvj.12atlas Work In Progress L = ~6.5 pb,.1.8.6.4.2! s = 13 TeV.2.4.6.8 1 1.2 lv P T /M lvj Shows optimal cut is D2 < 1.4 Shows optimal cut is P lv T/M lvj >.5 2
Trigger Studies 21 Old Trigger is passl1_em18vh and passl1_mu1 L1 = Level 1 Trigger New Trigger is passhlt_j36_a1_sub and passhlt_j46_a1_sub in addition to old trigger HLT = High Level Trigger a = AntiKt4LCTopo A1 = Large-R Jet 1 = size of cone 46 = P T threshold Sub = pile-up subtraction Subtracts interference with objects
22 Trigger Studies Large-R Jet mass between 65 GeV Large-R Jet mass between 65 GeV & D2 > 1.4 Large-R Jet mass between 65 GeV Large-R Jet mass between 65 GeV & D2 > 1.4
Multivariate Analysis 23 2 TeV targeted selection using Boosted Decision Trees Input Variables Large- R Jet Mass (FJMass) Lepton Neutrino P T (lvpt) Lepton Neutrino - Large-R Jet η (lvfjeta) Lepton Neutrino - Large-R Jet ϕ (lvfjphi) Large-R Jet P T (FJPt) Large-R Jet y (FJy) Large-R Jet D2 (FJD2)
Multivariate Analysis Input Variables 24
Multivariate Analysis cont. 25 Significance of 3.667 with optimal cut value of -.2 Large-R Jet P T is best discriminant against background
26 Multivariate Analysis cont. Cut Efficiencies Plot BDT Output Distributions and background efficiency for specific Boosted Decision Tree value Shows difference between signal and background due to their responses to the Boosted Decision Tree
Topology Plots 27 Show agreement between data and Monte Carlo No cuts made on D2 for plot of Large-R Jet D2 No cuts made on P lv T/M lvj for plot of P lv T/M lvj All other plots made using baseline cuts
28 Topology Plots cont. Electrons η 12 1 8 6 L = ~6.5 pb, s = 13 TeV # 25 2 15 L = ~6.5 pb, s = 13 TeV! Electrons P T 4 1 2 5 " 2.5 " 2 " 1.5 " 1 ".5.5 1 1.5 2 2.5 1 2 3 4 5 6 7 electrons! electrons P T Muons η 14 12 1 8 6 L = ~6.5 pb, s = 13 TeV # 2 18 16 14 12 1 8 L = ~6.5 pb, s = 13 TeV! Muons P T 4 2 6 4 2 " 2.5 " 2 " 1.5 " 1 ".5.5 1 1.5 2 2.5 1 2 3 4 5 6 7 8 91 muons! muons P T
Topology Plots cont. Large-R Jet η 4.5 4 3.5 3 2.5 2 1.5 1.5 L = ~6.5 pb, s = 13 TeV # " 2 " 1 1 2 J! 6 5 4 3 2 1 L = ~6.5 pb, s = 13 TeV! 2253354455556657 J P T [GeV] 29 Large-R Jet P T Large-R Jet D2 1 9 8 7 6 5 4 3 L = ~6.5 pb, s = 13 TeV! 25 2 15 1 L = ~6.5 pb, s = 13 TeV! Large-R Jet Mass 2 1 5.5 1 1.5 2 2.5 3 3.5 4 J D2 2 4 6 8 112141618 [GeV] M J
Topology Plots cont. Lepton Neutrino η 4 3.5 3 2.5 2 1.5 1.5 L = ~6.5 pb, s = 13 TeV # " 2.5 " 2 " 1.5 " 1 ".5.5 1 1.5 2 2.5 lv! 1 9 8 7 6 5 4 3 2 1 L = ~6.5 pb, s = 13 TeV! 3 4 5 6 7 8 9 1 lv P T [GeV] 3 Lepton Neutrino P T Lepton Neutrino Mass 6 5 4 3 2 1 L = ~6.5 pb, s = 13 TeV! 8 7 6 5 4 3 2 1 L = ~6.5 pb, s = 13 TeV! Lepton Neutrino & Large-R Jet P lv T /M lvj 4 5 6 7 8 9 11112 M vv [GeV].1.2.3.4.5.6.7.8 lvj P T /M lvj
Topology Plots cont. 31 Large-R Jet y Number of Large-R Jets 6 5 4 3 L = ~6.5 pb, s = 13 TeV " 2 18 16 14 12 1 L = ~6.5 pb, s = 13 TeV! 2 8 6 1 4 2! 2! 1 1 2.5 1 1.5 2 2.5 3 J y Number of J
Conclusions Cut based vs. Multivariate Analysis Multivariate analysis gives better sensitivity results More difficulo estimate systematic uncertainties using multivariate analysis approach Cut based is proposed method Final cuts Tight electron, medium muon L1 Trigger MET > 1 GeV Lepton Neutrino P T > 2 GeV At least 1 Large-R Jet with P T > 2 GeV Large-R Jet D2 < 1.4 P lv T/M lvj >.5 I would like to Poppy for her mentorship and the Columbia REU Program and Prof. Parsons for this opportunity! 32