Model Independent Search for Deviations from the Standard Model ...

More documents

Recommendations

Info

$Planungsraster FP C SoSe2013 Di. 13:30-18:00 h Gruppe \ Tag 9.4 ...$

102 Gaussians is the quadratic sum of both individual widths, this can be incorporated by additionally smearing the electrons and muons with a Gaussian of the width σ add : E ′ = E + E · Gauss(0, σmean) (8.11) E ′′ = E + E · Gauss(0, σup) (8.12) = E ′ + E ′ · Gauss(0, σ add ) (8.13) = E + E · Gauss(0, σmean) + E · Gauss(0, σ add ) + (8.14) E · Gauss(0, σmean) · Gauss(0, σ add ) } {{ } ≈0 =⇒ σadd 2 = up σ2 − σmean 2 . (8.15) Independent of this, the electron MET is smeared using Equation 5.7. Ultimately this results in two histograms for muons (mean and add-smeared), and three histograms for electrons (mean, add-smeared and MET-smeared). Implementing the error in the Search Algorithm is similar to the energy scale case. Again all bins and all MC are correlated as the smearing parameters hold for all leptons and all MET energy ranges of all MC. For the retrieval of the Region of Interest the smearing error is computed for each region seperately: σ 2 smear = ⎛ ⎝ ∑ bin i ∑ N j i (add-smeared) − ∑ MC j bin i MC j The error of MET in the electron case can be computed in the same way. ∑ ⎞2 N j i (mean) ⎠ . (8.16) Also in the test of General Signicance, special care has to be taken to preserve the direction of the smearing error throughout the whole distribution as all bins are statistically dependent. In contrast to the energy scale, the down error is not available. To account for possiblysmallersmearingparameters,theerrordeterminedfromthedierencebetweenthe add-smeared histogram and MC mean histogram is simply symmetrised. Again, a random number g can be diced once using a Gaussian with mean= 0 and σ = 1, where the sign xes the direction of the smearing error (up or down) and the absolute value determinines the magnitude of the error. For each bin a new entry can then be generated using: ⎛ bin new = ⎝ ∑ MC j ⎞ ⎛ ⎞ N j i (mean) ⎠+g · ⎝ ∑ N j i (add-smeared) − ∑ N j i (mean) ⎠ . (8.17) MC j MC j
8.6. Summary of the Different Contributions 103 8.6 Summary of the Dierent Contributions Some nal remarks concerning the dominant contributions to the systematic uncertainty should be made. As most event classes are very dierent from each other, the systematic error varies much from one class to another. In the case of the multiplicative factors, the magnitudes are easy to estimate, e.g for a single muon class : MC-correction-factor σ rel ≈ 3%, MC cross sectionuncertainty σ rel ≈ 3.5%as W-productiondomiantesanderroronQCD-background σ rel ≈ 40%. The other two systematic uncertainties (energy scale and smearing) are more dicult to gure as they shift the whole distribution. Again, the dominate contribution strongly dependsonthespeciceventclass: Inaclasswithjets,thesmearingerrorscanbeneglected with respect to the Jet Energy Scale uncertainty; in a MET distribution with electrons, the MET-smearing will be the dominant error, next to the scale factor uncertainties. To conclude, the systematic uncertainties are dierent for every event class and for every Region of Interest. In the event classes with huge statistics, 1l(+X) and 1l 1j(+X), all systematic errors combined are 10%. In the exotic classes and the high-p T tails of the distributions, they can also reach values of ≈ 50%, due to the fact that here smearing and energy scale uncertainty redistribute the bin entries (e.g. unsmeared bin entry: N MC = 4, smeared bin entry: N MC = 2).
Page 1:
Model Independent Search for Deviat
Page 4 and 5:
Contents 4 The Data Sample and Obje
Page 7 and 8:
Introduction Science attempts to de
Page 10 and 11:
4 The Theoretical Foundations 1.1.2
Page 12 and 13:
6 The Theoretical Foundations e - e
Page 14 and 15:
8 The Theoretical Foundations Figur
Page 16 and 17:
10 The Theoretical Foundations Figu
Page 18 and 19:
12 The Theoretical Foundations 1.3
Page 20 and 21:
14 The Theoretical Foundations cont
Page 22 and 23:
16 Tevatron and the DØ-Detector Fi
Page 24 and 25:
Page 26 and 27:
Page 28 and 29:
22 Tevatron and the DØ-Detector 1/
Page 30 and 31:
24 Tevatron and the DØ-Detector 2.
Page 32 and 33:
26 Tevatron and the DØ-Detector of
Page 35 and 36:
Chapter 3 The Concept of Model Inde
Page 37 and 38:
3.2. Concept 31 viations from SM. T
Page 39:
3.2. Concept 33 cal problem as many
Page 42 and 43:
36 The Data Sample and Object Ident
Page 44 and 45:
Page 46 and 47:
Page 48 and 49:
Page 50 and 51:
muon_trackchi_pt_py Entries 5727 mu
Page 52 and 53:
Page 54 and 55:
jet_eta_phi_px Entries 29768 Mean j
Page 57 and 58: Chapter 5 Monte Carlo Samples 5.1 M
Page 59 and 60: 5.1. MC Generator 53 Figure 5.3: Sc
Page 61 and 62: 5.2. SM-Processes 55 SM-process (M
Page 63 and 64: 5.3. Energy Scale and Smearing 57 J
Page 65: 5.3. Energy Scale and Smearing 59 e
Page 68 and 69: 62 General Data↔MonteCarlo Compar
Page 96 and 97: 90 Search Algorithm oer sucient sen
Page 98 and 99: 92 Search Algorithm probability 0.0
Page 100 and 101: 94 Search Algorithm 7.2.2 The Princ
Page 103 and 104: Chapter 8 Systematic Uncertainties
Page 105 and 106: 8.3. QCD-Background 99 Whendicingda
Page 107: 8.5. Smearing 101 Here N j i (σ+ )
Page 112 and 113: 106 Results and Interpretation Figu
Page 114 and 115: 108 Results and Interpretation Even
Page 116 and 117: 110 Results and Interpretation even
Page 118 and 119: 112 Results and Interpretation Anat
Page 120 and 121: 114 Results and Interpretation To c
Page 122 and 123: 116 Results and Interpretation Resu
Page 126 and 127: 120 Results and Interpretation Even
Page 130 and 131: 124 Results and Interpretation Resu
Page 133 and 134: Chapter 10 Conclusion In this diplo
Page 135 and 136: Appendix A Trigger Denitions Electr
Page 137: 131 Muon-triggers for lists up to v
Page 140 and 141: 134 Bibliography [19] A. Aktas et a
Page 143 and 144: Acknowledgements So, the end is nea
show all

Model Independent Search for Deviations from the Standard Model ...

Create successful ePaper yourself

Delete template?

Save as template?