Title: Mitchell Naisbit Manchester University Elba 2003
11.Aim and Definitions
- Aim is to formulate an efficient selection
procedure for p po decays of the t - Analysis will optimise
efficiency
purity
rej. factor
Nsnumber signal decays in MC sample (2NxBR),
nsnumber truth-matched candidates selected
btotal number of background decays selected
Nb number background decays in sample, nbnumber
background decays selected
- To combine MC will scale to a data sample of (so
far) 0.134 fb-1 of data
Mitchell Naisbit
Manchester University
Elba 2003
22.TauUser Preselection
Nano Cuts
Input 1M generic tt- events
Mitchell Naisbit
Elba 2003
Mitchell Naisbit
Manchester University
Elba 2003
32. TauUser Preselection
Nano Cuts
Input 1M generic tt- events
Level-3 trigger cut
Mitchell Naisbit
Elba 2003
42. TauUser Preselection
Nano Cuts
Input 1M generic tt- events
Level-3 trigger cut
Stream19
42 pass nanofilter
Mitchell Naisbit
Elba 2003
52.TauUser Preselection
Micro Cuts
Events input from nanofilter
Mitchell Naisbit
Elba 2003
62.TauUser Preselection
Micro Cuts
Events Input from nanofilter
tracks even and lt8
Mitchell Naisbit
Elba 2003
72.TauUser Preselection
Micro Cuts
Events Input from nanofilter
tracks even and lt8
Zero net charge
Mitchell Naisbit
Elba 2003
82.TauUser Preselection
Micro Cuts
Events Input from nanofilter
tracks even and lt8
Zero net charge
Topology classification
36 pass preselection
Mitchell Naisbit
Elba 2003
93.Selection Procedure
Selection for data and MC samples
For 1-1 events backgrounds reduced by tag
Selection requirements
p PID (piLHLoose)
p not in eMicroLoose
Signal side
Single AWG po
e or ยต PID as tag
Tag side
or
As above for ? tag
Mitchell Naisbit
Elba 2003
104.Monte Carlo truth matching
recon. p, po, ?A, ?B must match back to particle
in truth list of same identity
Reconstruction Requirements for signal t decays
p and po must have same mother which is a true ?
?A and ?B must have same mother which is a true
po
Mitchell Naisbit
Elba 2003
115.Results tt- MC
(1)
Running on 1M generic tt- events
Ns 508,200 signal decays
(1)
Truth codes for selected decays
(2)
(2)
Truth codes for truth-matched candidates
ns 48,613
(3)
Backgrounds
(3)
btau 21,401
Main t backgrounds are a1 decays and improperly
reconstructed signal decays.
Mitchell Naisbit
Elba 2003
125.Results background MC
Analysed MC samples of -
(hadronic)
(leptonic)
(photon)
Giving scaled results-
Ns61,385 ns5,872 bt2,585
bother2,208
b bt bother 4,793
Mitchell Naisbit
Elba 2003
135.Results
p 55.10
e 9.57
Main backgrounds
- Non-signal t (a1)
- Improperly reconstructed
- signal decays
- Leptonic (bhabhaFSR)
- and some hadronic
Mitchell Naisbit
Elba 2003
146.Event Parameter Distributions - tt- MC
nNeutrals
Ranges from 2 to 15
nChargedTracks
Ranges 0 to 6
thrustMag
nNeutrals
Ranges from 0.79 to 0.99
(but no quantities are of use to eliminate t)
Mitchell Naisbit
Elba 2003
156.Event Parameter Distributions - tt- MC
nNeutrals
Ranges from 2 to 15
nChargedTracks
Ranges 0 to 6
thrustMag
nNeutrals
Ranges from 0.79 to 0.99
(but no quantities are of use to eliminate t)
Mitchell Naisbit
Elba 2003
166.Event Parameter Distributions - tt- MC
nNeutrals
Ranges from 2 to 15
nChargedTracks
Ranges 0 to 6
thrustMag
nNeutrals
Ranges from 0.79 to 0.99
(but no quantities are of use to eliminate t)
Mitchell Naisbit
Elba 2003
176.Event Parameter Distributions - background MC
nNeutrals
nChargedTracks
cut gt 12 (hadronics)
cut all gt 6 (hadronics)
Mitchell Naisbit
Elba 2003
186.Event Parameter Distributions - background MC
cut lt 0.8 (hadronics)
cut gt 0.99 (leptonics)
thrustMag
Mitchell Naisbit
Elba 2003
197.Efficiency and Purity Optimisation
nNeutrals
nCharged Tracks
cut gt 6
cut gt 12
thrustMag
thrustMag
cut gt 0.988
cut lt 0.79
Mitchell Naisbit
Elba 2003
208.Invariant Mass Spectrum
e 9.57
e 9.51
p 55.10
p 65.52
r 100
Mitchell Naisbit
Elba 2003
219.Future Work
- Finalise Selection Procedure
- Complete study into t non-t backgrounds and how
- to minimise them
- Perform fits to the invariant mass spectrum
- Implement detector unfolding techniques
Mitchell Naisbit
Elba 2003