2009.05.04 LDA: More SMD info, 3x3 tower energy, correlation matrix

Cut optimization with Fisher's LDA classifier

ROOT implementation for LDA:

Application for cuts optimization in the gamma-jet analysis

LDA configuration: default

LDA input parameters Set0:

  1. Set4 (link for results with LDA Set0-Set3):
    • Energy fraction in 3x3 cluster within a r=0.7 radius:
      E_3x3/E_0.7
    • Photon-jet pt balance:
      [pt_gamma-pt_jet]/pt_gamma
    • Number of charge tracks within r=0.7 around gamma candidate:
      Ncharge
    • Number of Endcap towersL fired within r=0.7 around gamma candidate:
      NtowBarrel
    • Number of Barrel towers fired within r=0.7 around gamma candidate
      NtowEndcap
    • Shower shape analysis: distance to 80% cut line:
      distance to cut line
    • Energy fraction in E_2x1 and E_2x2 witin E_3x3:
      E_2x1/E_2x2 and E_2x2/E_3x3
    • Energy in post-shower layer under 3x3 tower patch:
      E_post^3x3
    • Tower energy in 3x3 patch:
      E_tow^3x3
    • SMD-u energy in 25 central strips:
      E_smd-u^25
    • SMD-v energy in 25 central strips:
      E_smd-v^25
    • SMD-v peak energy (in 5 central strips):
      E_peak

The number of strips in SMD u or v planes is required to be greater than 3

Pre-shower sorting (energy in tiles under 3x3 tower patch):

  1. pre1=0, pre2=0
  2. pre1=0, pre2>0
  3. 0 < pre1 < 0.004
  4. 0.004 < pre1 < 0.01
  5. pre1 < 0.01
  6. pre1 >= 0.01

Integrated yields per pre-shower bin:

sample total integral pre1=0,pre2=0 pre1=0, pre2>0 0 < pre1 < 0.004 0.004 < pre1 < 0.01 pre1 < 0.01 pre1 >= 0.01
photon-jet 2.5640e+03 3.5034e+02 5.2041e+02 5.6741e+02 5.2619e+02 1.9644e+03 5.9994e+02
QCD 5.6345e+04 1.3515e+03 4.3010e+03 1.2289e+04 1.5759e+04 3.3701e+04 2.2644e+04
pp2006 6.2811e+04 6.8000e+02 2.4310e+03 1.2195e+04 1.6766e+04 3.2072e+04 3.0739e+04

Photon pt and rapidity cuts:

  1. pt>7GeV
  2. pt>8GeV
  3. pt>9GeV
  4. pt>10GeV
  5. detector eta <1.4 (pt>7GeV)
  6. detector eta > 1.4 (pt>7GeV)

LDA Set4

Figure 1: LDA discriminant with Set0: Data to Monte-Carlo comparison (pt>7GeV cut)

Right plot for each pre-shower condition shows the ratio of pp2006 data to sum of the Monte-Carlo samples
Colour coding:
black pp2006 data, red gamma-jet MC, green QCD MC, blue gamma-jet+QCD


Figure 2: rejection vs. efficiency

Figure 3: purity vs. efficiency

Figure 4: purity vs. rejection

Figure 5: Correlation matrix (pt>7GeV cut)
pre1=0, pre2=0

pre1=0, pre2>0

0 < pre1 < 0.004

0.004 < pre1 < 0.01

pre1 < 0.01

pre1 >= 0.01