Transcript ppt - University College London

ttH, H->bb: Update
Janice Drohan
University College London
Overview:
Atlfast Samples
Top reconstruction: Χ2 vs Δ2
UK Atlas Higgs Meeting - RHUL - 22 April 2005
1
ttH, H->bb: Update
 Generation Status:
Sample
Generator
# Atlfast AOD Events
Restricted decay
of tops
Pythia
300,000
Yes
Jimmy
55,000
Yes
Pythia
22M
No
mc@nlo
720,000
No
qq->ttbb
AcerMC
400,000
No
gg->ttbb
AcerMC
1.8M
No
ttH 120GeV
ttjj
 Signal: Pythia vs Herwig (Jimmy)
 Very good agreement, slightly more jets in events
 Problems:
 Producing large numbers of mc@nlo events.
 Still to do:
 Need to produce a ttjj (light jets only) sample

Cammin produced this by filtering at Atlfast level
UK Atlas Higgs Meeting - RHUL - 22 April 2005
2
Top Reconstruction
 TDR Method:
 Minimises: Δ2=(mlvb-mt)2+(mjjb-mt)2
 Correct Jets assigned to Higgs ~25% cases
 Started looking at 2nd lowest Δ2 value:
 Produces
correct Higgs
in a similar
number of
cases
MH (MeV)
UK Atlas Higgs Meeting - RHUL - 22 April 2005
3
Top Reconstruction
MH (MeV)
UK Atlas Higgs Meeting - RHUL - 22 April 2005
4
Top Reconstruction
 Trying to use instead of Δ2:
2
2
2
(
m

m
)
(
m

m
)
(
m

m
)
 2  jjb 2 t  lvb 2 t  jj 2 W
 jjb
 lvb
 jj
 Using errors from within Atlfast Code for Jet
energies and electron energy:
Jet:

0.5

 0.03
E
E

0.245 0.12
 0.007
e: E  p 
E
T
eta<1.4
 Don’t yet have muon energy error
 S. Dean writing helper function to do this.
UK Atlas Higgs Meeting - RHUL - 22 April 2005
5
Top Reconstruction
 Errors looking reasonable:
mjjb
σmjjb
mlvb
σmlvb
mjj
σmjj
152.9
10.5
270.2
12.5
105.1
11.4
307.8
17.1
138.5
24.3
65.4
33.5
 Plans:
 Look into performance of χ2
 Assigning of b-jets to Higgs in case of 3 left
 138 events out of 2155, 60 wrong combinations
when choosing jets with highest pT
 ANN for signal/background separation
UK Atlas Higgs Meeting - RHUL - 22 April 2005
6