Документ взят из кэша поисковой машины. Адрес оригинального документа : http://qfthep.sinp.msu.ru/talks2015/QFTHEP%20-%20Myagkov.pdf Дата изменения: Fri Jun 26 13:53:26 2015 Дата индексирования: Sat Apr 9 23:25:14 2016 Кодировка:

Limits on FCNC couplings in single top events
I G O R M YA G K O V S I N P M S U , O N B E H A L F O F CMS C O L L A B O R AT I O N
6/26/2015

1


FCNC vertex in single top: tug & tcg
Wbj channel (representative diagrams):

· Monte-Carlo generated with CompHEP: = 0.03
-1

= 0.03

-1

· All necessary diagrams

6/26/2015

2


Strategy
Setup
Basic selection Reweighting
Preselected events

QCD BNN

FCNC input vars

FCNC BNN

BNN output

Statistical analysis

FCNC Limits

Published as CMS-PAS-TOP-14-007

6/26/2015

3


Analysis setup, event selection & reweighting
Setup: · Int. luminosity 5 -1 @ 7 · Only channel considered
Selection:
Object definitions follow CMS Top Group recommendation

Triggers: · HLT_IsoMu17_v* · HLT_IsoMu24_v* · HLT_IsoMu24_eta2p1_v*
MC reweighting: · Pile-Up Standard reweighting based on "true" number of interaction · B-tagging Recommended Rizzi recipe Scale factors provided by BtagPOG · Triggers and muon ID/Iso 8 TeV SF provided by MuonPOG 7 TeV SF measured by working group · PDF
4

· Quality criteria for primary vertex · Only one "tight " lepton:
: > 26 : > 10 J: > 30


, < 2.1, < 0.12 , < 2.4, < 0.2
(1,2)

· No more additional "loose" leptons:


· Two or three Jets:


, < 4.7,

> 40





· At least one b-tagged jet according to CSVT · At least one untagged jet according to CSVT
6/26/2015


Systematics
Sources of systematic uncertainties:
· Finite MC statistics · Luminosity

Marginalized:
· · · · · · · · · Xsections JEC JER Unclustered MET PileUp B-tag /mistag Triggers SF Lepton Id Lepton Iso

Unmarginalized:
· · · · Generator choise Scale Matching PDF

Unmarginalized uncertaintites are estimated with toys. Pseudodata are constructed with a best-fit value for t-channel x-section, not SM value.

6/26/2015

5


Bayesian Neural Networks
QCD suppression
Estimation
Multijet QCD background is estimated from data Modified event selection "Cleaning" procedure for jets

Suppression
Special BNN for QCD removal BNN output > 0.7

6/26/2015

6


FCNC limits
BNNs trained to separate tug and tcg processes from SM. Input variables: · Optimal variables method: Feynman diagram structure analysis.

Signal:
· Events with FCNC Background: · All SM processes

Scenarios:
· tug ­ 1-dimensional, · tcg ­ 1-dimensional,


=0 =0

· tug + tcg ­ 2-dimensional

6/26/2015

7


One-dimensional tug scenario
Observed (expected) 1D limit @ 95% C.L.:
·


< 1.8 10

-2

1.2 10

-2



-1

In terms of branching fraction:
· + < 3.55 10-4 (1.58 10-4 )

6/26/2015

8


One-dimensional tcg scenario
Observed (expected) 1D limit @ 95% C.L.:
·


< 5.6 10-2 3.1 10-2

-1

In terms of branching fraction:
· + < 3.44 10
-4

(1.05 10-4 )

6/26/2015

9


Two-dimensional scenario

6/26/2015

10


Alternative analyzes
FCNC in single top in association with Z @ 7TeV: CMS PAS TOP-12-021

6/26/2015

11


Alternative analyzes
Single top in association with : CMS PAS TOP-14-003

6/26/2015

12


Summary
· Obtained 1D limits on


and



FCNC couplings

· As well as 2D limits from simultaneous fit
· CMS-PAS-TOP-14-007 is published

· Updated results with 7+8 TeV full datasets are being prepared for publication as CMS-TOP-14-007 paper

Thank you for your attention!

6/26/2015

13