Документ взят из кэша поисковой машины. Адрес оригинального документа : http://qfthep.sinp.msu.ru/talks2015/QFTHEP15_Vorotnikov_v9.pdf Дата изменения: Fri Jun 26 12:31:04 2015 Дата индексирования: Sat Apr 9 23:28:24 2016 Кодировка:

Limits on anomalous Wtb couplings in single top-quark events in the CMS experiment

Georgii Vorotnikov SINP MSU on behalf of the CMS Collaboration

QFTHEP'15 26 June 2015


Anomalous coupling in Wtb vertex
General form of the effective Wtb vertex lagrangian:

2

Where

q'
q
+

f ,f fT ,f
L

L V

R V R T

­ Left and right vector couplings ­ Left and right tensor couplings

W
+

t
b

W

+


b

SM values: f

L V R V

= 1; = fT = f
L R T

g

b

f

=0


Measurement of the W boson helicity in events with a single reconstructed top quark in pp collisions at sqrt(s) = 8 TeV
Setup Preselected events
Top quark reconstruction Anomalous coupling limits

3

Basic selection

Reweighting

Statistical analysis

W helicity fractions fits

Published as CMS-TOP-12-020, JHEP01(2015)053


Analysis setup
Setup: Int. luminosity 19.7 /fb at

4

s =8 TeV


MC Reweighting: Pile-Up B -ta g Triggers and lepton ID/Iso

Selection: Objects definition and selection follows CMS Top Group recommendations; e and channels






Two jets with one b-tagged and one untagged jet according to TCHPT R(l,jet)>0.3 cut for jets
m W > 50 GeV/c T





Triggers: HLT_I soEle 27_v* HLT_I soMu2 4_eta 2p1_ v*


W boson helicity fractions
Expected distribution of the partial width of top quark decay:

5

Where

F L , F0 , F

R

­ Left, longitudinal and right polarization fraction of the W-boson helicity
q'
q

NNLO SM values: F L= 0.311± 0.005 ,

F 0= 0.687 ± 0.005 , F R = 0.0017 ± 0.0001
­ angle in top quark rest frame between the lepton 3-momentum in W-boson rest frame and the 3momentum of W-boson

W
+

t
b

W

+


b

+

g b


Top quark reconstruction
To calculate , top quark 4-momenta is needed. For this purpose neutrino longitudinal momentum will be found.

6

Where Events with complex p
z ,

, are excluded from the analysis


The fit method
The Poisson likelihood function

7

Where

n

data i

­ number of selected data events in the bin i ­ expected nubmer of simulated events



MC ; F i



signal i

,

Wj

f
ets

is the shapes of singnal, W+jets and other backgrounds sources and are defined from simulation, is signal normalization parameter (fixed to 1) and is free parameter for normalization W+jets background

Wjets i

and

bkg - other i

Taking into accout



F i =1 , there are only three parameters in likehood function


Systematics
Sources of systematic uncertainties: Experimental


8

Modelling


Normalization


Method-specific:


JEC / JER Unclustered MET PileUp B-tag / mistag Triggers SF Lepton Id / Iso Luminosity

Generator choise Scale Top quark mass P DF W+jets shape uncertainty

ttbar Single t quark QCD multijet Electroweak backgrounds

SM W helicities in MC Fixing signal normalization in the fit Finite MC statistic Wtb vertex bias from anomalous couplings


































Standard Model results
Muon and electron channels combination:

9

SM values (NNLO):

F L =0.311 ± 0.005 , Czarnecki at al. F 0 = 0.687 ± 0.005 , F R = 0.0017 ± 0.0001


Anomalous couplings in W helicity fractions
The anomalous couplings can be obtained from partial width for the top decay into a W boson with -1, 0 or +1 helicity:
i= W i

10

Where

f f

L, R V L,R T

= Re ( V = Re ( g

L,R L, R

) )

If CP is conserved, the couplings could be taken as real


Anomalous Wtb results
f
R T

11

f

R T

CMS, W helicity in ttbar, 7 TeV

f

L T

f

R T

Best fit:
L f T =-0.017 , R f T =-0.008

f

L T

LHC combination, W-helicity, 7 TeV

f

L T


Search for anomalous Wtb couplings and FCNC in t-channel single-top-quark events
Setup Preselected events
SM input vars BNN output

12

Basic selection

Reweighting

QCD BNN survived events

SM BNN

SM Stat. analysis measurements

QCD BNN

Wtb input vars

Wtb BNNs

BNN output

Stat. analysis

Wtb
measurements

Published as CMS-PAS-TOP-14-007


Analysis setup
Setup: Int. luminosity 5 / fb at

13

s =7 TeV


MC Reweighting: Pile-Up B -ta g Triggers and muon ID/Iso P DF

Selection: Objects definition and selection follows CMS Top Group recommendations;






muon channel only Two or three jets At least one b-tagged jet according to CSVT, at least one untagged Triggers: HLT _Iso Mu17_ v* HLT _Iso Mu24_ v* HLT _Iso Mu24_ eta2p 1_v*






Multijet QCD suppression


14

E s ti m a ti o n Multijet QCD background is estimated from the d a ta Modified event selection ''Cleaning'' procedure for jets, R(l,jet)>0.5 or removing jet Suppression:









Special neural network for QCD removing BNN_qcd > 0.7 cut




High level analysis
Input variables : Optimal variables method: Feynman diagrams structure analysis Signal: t-channel production (SM analysis) Background: All other processes (SM analysis) or left vector coupling ( f V )
L
R V L T

15

or

one of f

or f

coupling (Wtb analysis)

(Wtb analysis)


Systematics
Sources of systematic uncertainties:


16

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 P DF Unmarginalized uncertaintites are estimated with toy experiments. Pseudodata are constructed with a best-fit value for tchannel x-section, not SM value.


























Standard Model results

17



observed t - channel

= 69.74

+ 9.1 (+ 13.0 % ) -8.9 (- 12.8 % )

pb



theory , NLO t - channel

+ =65.9-2.1 0.7

+ 1.5 - 1.7

pb

(Kidonakis)



t - ch.

=67.2 ±6.1 pb

(JHEP12(2012)035)

Measured cross-section is used in anomalous Wtb couplings searches


Anomalous coupling searching

18

For f or f anomalous coupling additional BNN is trained to separate it from f (i.e. from Standard Model t-channel events)

R V

L T

L V

2 different scenarios:


L R L R ( f V , f V ) , while LT and RT couplings are fixed to SM values ( f T = f T = 0 ) Searching for



Searching for ( f V , f T ) , while RV and RT couplings are fixed to SM values( f V = f T = 0 ) Each scenario gives 1D limit for each coupling and 2D fit of couplings distribution

L

L

R

R


(LV,RV) scenario

19

Observed (expected) 1D limits at 95% C.L.:


(LV,LT) scenario

20

Observed (expected) 1D limits at 95% C.L.:


Anomalous Wtb searches results

21

L f V> 0.90 , R f V< 0.34 , L f T< 0.09

R f V2 > L f T2 < R f T2 <

0.89 , 0.07 , 0.18

D0, single top only, 1,96 TeV


Anomalous Wtb searches results
f
R T

22

f

R T

CMS-TOP-12-020

f

L T

L f V> 0.90 , R f V< 0.34 , L f T< 0.09

f

L T

f R =-0.008 ± 0.024 ( stat. ) T

+ 0.029 -0.030

( syst. )

CMS, W helicity in ttbar, 7 TeV


Summary


23

W-boson helicity is measured for 8 TeV in e- and - channels simultaneously 1D limits for LV, RV and LT couplings in Wtb vertex 2D limits for (LV,RV), (LV,LT) and (LT,RT) CMS-12-020 is already published in JHEP CMS-PAS-TOP-14-007 is published as a Physics Analysis Summary, updated results with 7+8 TeV full datasets, limit for RT coupling, ''3D'' scenarios will be published soon as CMS-TOP-14-007 paper









Thank you for attention!


Backup


Artificial Neural Networks (ANN)
ANN analysis:


25

Possibility for multi-dimentional analysis High efficiency Supervised learning
Neurons





Single-layer perceptron:
Connections

Bayesian Neuron Networks (BNN):


Bayesian treatment of probability: output is interpreted as degree of belief



Ansambles of networks: overfitting protection

Input layer Hidden layer

Output layer