Associated Production for the Standard Model Higgs at CDF

1
Associated Production for the Standard Model
Higgs at CDF

• D. Bortoletto
• Purdue University

• Outline
• Higgs at the Tevatron
• WH? ?? bb
• ZH ? ?? bb
• WH and ZH ? met bb
• Outlook

2
SM Higgs

• The Higgs remains the Holy Grail of Particle
  Physics
• Direct LEP searches MHgt114 GeV at 95 CL
• Indirect searches improving. Smaller error on top
  and W mass Mtop170.9 1.8 GeV and MW80.398
  0.25 GeV

MHlt144 GeV (ignoring direct limit) MHlt182 GeV
(including 114 GeV limit)
3
Higgs at the Tevatron

• mHlt135 GeV use
  W/Z H and H?bb
• Trigger on
  leptons or met
• mHgt135 GeV direct production and H?WW

Total
H ? WW?l ? l ?
WH ? l ? bb
ZH ???bb
ZH ? llbb
4
ZH???bb

• Strategy
• Optimize lepton selection to improve Z???
  acceptance by 70
• Improve S/B
• Two loose b-tag 50 efficient, 1.5 fake rate
• One tight b-tag 40 efficient, 0.5 fake rate
• Improve mass resolution
• Use MET to improve mass resolution using NN since
  no real MET expected

CDF secondary vertex
5
ZH???bb
(1,1)

• Reduce the background with 2D NN
• Training on ZZ, Zbb, tt and ZH MC
• Fakes and mistags from data

tt
fakes
tt versus ZH
Zjets
ZZ
ZH
(1,0)
(0,0)
mistags
Zjets versus ZH
40
6
Signal region

• Double tag

• Single tag

Expected 101.6 17.8 Data 100 events
Expected 12.8 3.5 Data 11 events
S/B 1/4 (before selections
5/1015)
7
Limits
mH115 GeV
Major systematic uncertainties Z jet
Background shape Z heavy flavor Jet Energy
scale Total systematic shift in ?? 0.19 pb
95 CL s over SM expectation
8
WH???bb
Observable
Lepton(e/µ) PTgt20 GeV, ?lt1
Missing ET gt20 GeV

• Backgrounds
• Non-W QCD (lepton fakes) Data based estimation
• Mistag Data based estimation
• WHeavy Flavor (Wbb, Wcc, Wc) Data and MC based
  estimation
• Other standard model background (tt, single top,
  di-boson, Z?tau tau)
• MC based estimation

9
NN Tagging

• Apply neural net b-tagger on the information of
  SECVTX to improve the separation between b-jets
  and c-jets or light-jets
• NN input Lxy significance, vertex mass,
  pseudo-ct

Keeping 90 of true b-jets, 65 of l-jets and 50
of c-jets are removed!
10
Results WH???bb
2 SECVTX without NN
At least 2 b-tagging (double tag)
NN b-tagging is NOT applied
Higgs
Expected 44.2 8.5 Observed 39
11
Limits
About 10 SM
12
W/Z?MET and H?bb

• This final state can
  be used to search for
• Signal has a distinctive
  topology
• Two jets
• ?lt2.0 and one ?lt0.9
• ET,jet1 gt 35 GeV
• ET,jet2 gt 25 GeV
• ?R(j1,j2)gt1.0
• No other jets with ETgt 20 GeV, ?lt2.0
• Large MET (gt50 GeV)
• b-jets (use b-tagging)

13
Z/W?MeT and H?bb

• Control Region 2 EWK
• 1 identified lepton
• ?F(j2, MET)gt0.4

• Unbiased analysis

• Control Region 1 QCD
• Veto events with identified leptons
• ?F(j2, MET)lt0.4

• Extended Signal Region
• Veto events with leptons
• ?F(j2, MET)gt0.4

14
Results

• Sensitivity Optimization performed in extended
  signal region
• ET,1 gt 60 GeV and ?(ET,1, MET) gt 0.8
• MET gt 70 GeV
• HT scalar sum of the jet Et-s
• Missing HT (vectorial sum of jet Et-s)
• Missing HT / HT gt 0.45

Tight SECVTX applied to this analysis both for
single and double tagging
Double tag
Expect 14.8 2.65 Observe 16
Double b- tag CR2 Dominated by top
15
Exclusion limits
Systematic uncertainties B-tagging 8.6 for
double tags Jet energy scale 7-26 Luminosity
6 Z/W heavy flavor normalization 40
16
Candidate event
Double-tagged event
Missing ET 144.8 GeV
Second Jet ET 54.7 GeV 
Leading Jet ET 100.3 GeV
Di-jet invariant mass 82 GeV
17
Combination

• Combination of all CDF channels and with D0 in
  ongoing

CDF Combination does not include the improved
llbb analysis
ICHEP 06
18
Outlook

• Current limits with 1 fb-1
• 2.5 fb-1 have been collected

• Tevatron operating well ? 6 fb-1 by 09

2007

• Improved analysis methods

• Improved triggers implemented
• Ex. New Calorimeter trigger
• We keep working and searching

19

• Backup slides

20
Validation

• Pre-tagged sample

21
Results WH???bb
1 SECVTX with NN tagging
Expected 394.4 66.6 Observed 421
22
Backgrounds
A few examples!
Tight SECVTX applied to this analysis both for
single and double tagging
Double b- tag CR2 Dominated by top
Single tag CR1 Dominated by QCD and mistags
23
Signal Region
S/B Optimization is performed in extended signal
region Jet energy Missing ET HT scalar sum of
the jet Et-s Missing HT (vectorial sum of jet
Et-s)

• ?(1st Jet, Missing ET) gt 0.8
• Missing ET gt 70 GeV
• Missing HT / HT gt 0.45
• 1st Jet ET gt 60 GeV

Single tag
Double tag
Expect 14.8 2.65 Observe 16
Expect 251.0 43.2 Observe 268