[PPT] - with new hadronic top-tagging techniques T T.A. du Pree, P. Harris, PowerPoint Presentation

SLIDE 1

Search for top+Emiss

T

with new hadronic top-tagging techniques

T.A. du Pree, P. Harris, J. Marrouche, N. Wardle [CERN]

M. Cremonesi, B. Jayatilaka, J. Lewis, C.M. Suarez, N. Tran [Fermilab]
D. Abercrombie, B. Allen, Z. Demiragli, G. G´
mez-Ceballos, D. Hsu,
Y. Iiyama, D. Kovalskyi, B. Maier, S. Narayanan, C. Paus [MIT]
K. Hahn, S. Sevova, K. Sung, M. Trovato [Northwestern]
J. Pazzini, M. Zanetti, A. Zucchetta [Padova]

MET+X, 24/02/2017

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 1 / 28

SLIDE 2

Analysis status

◮ Shown today: Summer16 MC and re-reco data ◮ AN and PAS will be updated in coming days with results presented today ◮ Early next week: first look at re-MINIAOD, updated JEC ◮ This talk: focus on updates since previous talk

◮ NLO signal model ◮ POG b-tag scale factors ◮ Additional data-driven backgrounds ◮ Categorization and optimization of selection

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 2 / 28

SLIDE 3

Monotop via flavor-mixing neutral current

Vector field that couples to q ¯ q′ and decays to χ¯ χ: Lint = Vµ ¯ χγµ(gVχ + gAχγ5)χ + ¯ quγµ(gVu + gAuγ5)quVµ + ¯ qdγµ(gVd + gAdγ5)qdVµ

u V g ui t ¯ χ χ

Definitions:

◮ Vµ is the spin1 mediator ◮ gVχ and gAχ are the vector and axial-vector couplings of the DM to the mediator ◮ gVu and gAu are 3 × 3 flavor matrices and are the vector and axial-vector couplings of the

up-type quarks (qu) to V

◮ gVd and gAd are the equivalent for down-type quarks (qd)

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 3 / 28

SLIDE 4

Monotop via flavor-mixing neutral current

◮ This is an extension of DMSimp with flavor mixing implemented

◮ CMS and ATLAS have agreed to use this model for monotop

◮ SU(2)L mandates gVu − gAu = gVd − gAd

◮ Choice: gVu = gVd and gAu = gAd

◮ We choose to only turn on u-t mixing in gV,Au ⇒ monotop

(GeV)

T

fat jet p 200 400 600 800 1000

Arbitrary units

0.1 0.2 0.3 = 0.3 TeV

V

m = 0.5 TeV

V

m = 1.0 TeV

V

m = 1.5 TeV

V

m = 2.25 TeV

V

m Simulation Preliminary

CMS

13 TeV

= 1 GeV

χ

= 1, m

V DM

= 0.25, g

V q

g

#AK4 jets 5 10

Arbitrary units

0.1 0.2 0.3 0.4 = 0.3 TeV

V

m = 0.5 TeV

V

m = 1.0 TeV

V

m = 1.5 TeV

V

m = 2.25 TeV

V

m Simulation Preliminary

CMS

13 TeV

= 1 GeV

χ

= 1, m

V DM

= 0.25, g

V q

g

PF MET (GeV) 200 400 600 800 1000

Arbitrary units

0.1 0.2 0.3 = 0.3 TeV

V

m = 0.5 TeV

V

m = 1.0 TeV

V

m = 1.5 TeV

V

m = 2.25 TeV

V

m Simulation Preliminary

CMS

13 TeV

= 1 GeV

χ

= 1, m

V DM

= 0.25, g

V q

g

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 4 / 28

SLIDE 5

Overview of objects

Jets

◮ AK4 jets for background suppression

◮ b-tagged if passes CSVL ◮ “Isolated” if no overlap with CA15 jet

◮ PUPPI CA15 jets to identify

hadronically-decaying tops

◮ b-tagged if one subjet passes CSVL ◮ Loose (tight) top-tagged if BDT score is

greater than 0.1 (0.45)

◮ Require 110 < mSD < 210 GeV

Tau veto

◮ Decay mode finding and POG very loose

isolation Electrons, muons, photons

◮ Latest POG IDs used for veto (SR) and

selection (CRs)

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 5 / 28

SLIDE 6

Selection overview

Preselection:

◮ Exactly one fatjet passing mass and BDT cuts ◮ Recoil greater than 250 GeV ◮ No identified τ leptons

Region(s) Main process(es) Ne/µ Nγ Niso

b-tag

Fatjet b-tag Signal Z → νν, W → (ℓ)ν, t¯ t → bqq′ + b(ℓ)ν >CSVL Single-ℓ (top) t¯ t → bqq′ + bℓν 1 1 >CSVL Single-ℓ (W) W → ℓν 1 <CSVL Dilepton Z → ℓℓ 2 − − Photon γ 1 − − NB: muon and electron regions are separate

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 6 / 28

SLIDE 7

Top-tagger

◮ Using a new BDT constructed out of substructure variables (particularly energy

correlation functions)

◮ Documentation available here, currently under review by JMAR

1 0.8 0.6 0.4 0.2 0.2 0.4 0.6 0.8 1

Events

200 400 600 800 1000 1200 1400 1600 1800 2000 2200 Data Z+jets t t W+jets Single t Diboson QCD Data Z+jets t t W+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb Top BDT 1 − 0.8 − 0.6 − 0.4 − 0.2 − 0.2 0.4 0.6 0.8 1

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

Dimuon selection: good agreement

1 0.8 0.6 0.4 0.2 0.2 0.4 0.6 0.8 1

Events

500 1000 1500 2000 2500 Data t t W+jets Z+jets Single t Diboson QCD Data t t W+jets Z+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb Top BDT 1 − 0.8 − 0.6 − 0.4 − 0.2 − 0.2 0.4 0.6 0.8 1

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

Muon t¯ t selection: good agreement in shape

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 7 / 28

SLIDE 8

Brief aside: t¯ t modeling

◮ Our nominal simulation (Powheg) does not get the t¯

t normalization correct

◮ Observed in other analyses that see high pT t¯

t

◮ Ongoing studies to understand and fix the problem, see talk at TopModGen ◮ Attempting to request a large FXFX sample to use in the analysis

Powheg

60 80 100 120 140 160 180 200 220 240

Events

1000 2000 3000 4000 5000 Data t t W+jets Z+jets Single t Diboson QCD Data t t W+jets Z+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb [GeV]

SD

fatjet m 60 80 100 120 140 160 180 200 220 240

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

amc@nlo FXFX

60 80 100 120 140 160 180 200 220 240

Events

500 1000 1500 2000 2500 3000 3500 4000 4500 Data t t W+jets Z+jets Single t Diboson QCD Data t t W+jets Z+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb [GeV]

SD

fatjet m 60 80 100 120 140 160 180 200 220 240

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

Herwig

60 80 100 120 140 160 180 200 220 240

Events

500 1000 1500 2000 2500 3000 3500 4000 Data t t W+jets Z+jets Single t Diboson QCD Data t t W+jets Z+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb [GeV]

SD

fatjet m 60 80 100 120 140 160 180 200 220 240

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 8 / 28

SLIDE 9

Top-tagging in signal events

[GeV]

SD

fatjet m

50 100 150 200 250 300 350 400

a.u.

0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18

=2.5 TeV

V

m =1.5 TeV

V

m =1.0 TeV

V

m =0.5 TeV

V

m =0.3 TeV

V

m =2.5 TeV

V

m =1.5 TeV

V

m =1.0 TeV

V

m =0.5 TeV

V

m =0.3 TeV

V

m

CMSPreliminary

(13 TeV)

1

36.6 fb =1

χ V

=0.25, g

q V

=1 GeV, g

χ

m

Top BDT

1 − 0.8 − 0.6 − 0.4 − 0.2 − 0.2 0.4 0.6 0.8 1

a.u.

0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08

=2.5 TeV

V

m =1.5 TeV

V

m =1.0 TeV

V

m =0.5 TeV

V

m =0.3 TeV

V

m =2.5 TeV

V

m =1.5 TeV

V

m =1.0 TeV

V

m =0.5 TeV

V

m =0.3 TeV

V

m

CMSPreliminary

(13 TeV)

1

36.6 fb =1

χ V

=0.25, g

q V

=1 GeV, g

χ

m

Generally observe higher mV ⇒ higher pt

T ⇒ more merged top jets

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 9 / 28

SLIDE 10

Signal region

◮ Tight category has much

higher t¯ t contribution than previous iterations of this analysis Loose

PF MET [GeV]

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10

=1 GeV

χ

=1.75 TeV, m

V

m Z+jets t t W+jets Single t Diboson QCD =1 GeV

χ

=1.75 TeV, m

V

m Z+jets t t W+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb

Tight

PF MET [GeV]

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10

=1 GeV

χ

=1.75 TeV, m

V

m Z+jets t t W+jets Single t Diboson QCD =1 GeV

χ

=1.75 TeV, m

V

m Z+jets t t W+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 10 / 28

SLIDE 11

Single µ (W) control region

◮ Prefit agreement is

reasonable

◮ One slightly suspicious bin

at 500 GeV

◮ NB: large t¯

t contamination in tight category Loose

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10 Data W+jets t t Z+jets Single t Diboson QCD Data W+jets t t Z+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb ) [GeV] µ PF U( 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

Tight

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10 Data W+jets t t Z+jets Single t Diboson QCD Data W+jets t t Z+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb ) [GeV] µ PF U( 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 11 / 28

SLIDE 12

Single µ (top) control region

◮ Normalization discrepancy

bserved in t¯

t

◮ Still investigating cause

and solution

◮ Some hints that the

FxFx simulation is better

◮ However, not a

show-stopper

◮ Will see that the fit

fixes this

Loose

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10 Data t t W+jets Z+jets Single t Diboson QCD Data t t W+jets Z+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb ) [GeV] µ PF U( 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

Tight

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10 Data t t W+jets Z+jets Single t Diboson QCD Data t t W+jets Z+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb ) [GeV] µ PF U( 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 12 / 28

SLIDE 13

Single e (W) control region

◮ Normalization discrepancy

bserved in all electron

regions

◮ Expect to go away with

updated samples Loose

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10 Data W+jets t t Z+jets Single t Diboson QCD Data W+jets t t Z+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb PF U(e) [GeV] 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

Tight

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10 Data W+jets t t Z+jets Single t Diboson QCD Data W+jets t t Z+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb PF U(e) [GeV] 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 13 / 28

SLIDE 14

Single e (top) control region

◮ Combination of electron

and t¯ t issues make this region look especially bad Loose

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10 Data t t W+jets Z+jets Single t Diboson QCD Data t t W+jets Z+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb PF U(e) [GeV] 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

Tight

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10 Data t t W+jets Z+jets Single t Diboson QCD Data t t W+jets Z+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb PF U(e) [GeV] 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 14 / 28

SLIDE 15

Dimuon control region

◮ Reasonable agreement

within large statistical uncertainties

◮ Some bin migration (data

events from loose to tight) in the 300-400 GeV range Loose

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10 Data Z+jets t t W+jets Single t Diboson QCD Data Z+jets t t W+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb ) [GeV] µ µ PF U( 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

Tight

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10 Data Z+jets t t W+jets Single t Diboson QCD Data Z+jets t t W+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb ) [GeV] µ µ PF U( 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 15 / 28

SLIDE 16

Dielectron control region

◮ Electron normalization

effect is observed once again Loose

300 400 500 600 700 800 900 1000

Events/GeV

3 −

10

2 −

10

1 −

10 1 10

2

10 Data Z+jets t t W+jets Single t Diboson QCD Data Z+jets t t W+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb PF U(ee) [GeV] 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

Tight

300 400 500 600 700 800 900 1000

Events/GeV

4 −

10

3 −

10

2 −

10

1 −

10 1 10

2

10 Data Z+jets t t W+jets Single t Diboson QCD Data Z+jets t t W+jets Single t Diboson QCD

CMSPreliminary

(13 TeV)

1

36.6 fb PF U(ee) [GeV] 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 16 / 28

SLIDE 17

Photon control region

◮ Slight excess observed in

low recoil on both categories

◮ Large statistical

uncertainties on MC Loose

300 400 500 600 700 800 900 1000

Events/GeV

2 −

10

1 −

10 1 10

2

10

3

10

4

10

Data +jets γ QCD Data +jets γ QCD

CMSPreliminary

(13 TeV)

1

36.6 fb ) [GeV] γ PF U( 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

Tight

300 400 500 600 700 800 900 1000

Events/GeV

2 −

10

1 −

10 1 10

2

10

3

10

Data +jets γ QCD Data +jets γ QCD

CMSPreliminary

(13 TeV)

1

36.6 fb ) [GeV] γ PF U( 300 400 500 600 700 800 900 1000

Exp Data-Exp 0.4 − 0.2 − 0.2 0.4

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 17 / 28

SLIDE 18

Fit strategy

As with previous iterations, the following data-driven estimations are performed:

◮ Z → ℓℓ and γ to constrain Z → νν ◮ W → ℓν to constrain W → (ℓ)ν ◮ t¯

t → bqq′ + bℓν to constrain t¯ t → bqq′ + (bℓ)ν In addition, we introduce two new constraints:

◮ Recall: tighter top-tag increases t¯

t contamination in W+jets control region

◮ Fix: use the t¯

t control region to constrain t¯ t in signal region and W+jets control region

◮ Technically accomplished by adding a link N W CR

t¯ t

↔ N SR

t¯ t

◮ The W/Z constraint (a l`

a monojet) is also used

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 18 / 28

SLIDE 19

Updated fit strategy

Z → νν SR Z → ℓℓ CR γ CR W → ℓν SR W → ℓν W CR t¯ t SR t¯ t W CR t¯ t top CR Solid links are new

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 19 / 28

SLIDE 20

Categorization and optimization

◮ Loose category is defined by BDT > 0.1

◮ Similar to ICHEP working point

◮ Tight category cut optimized on expected

limit

◮ Use Asimov data in all regions ◮ Signal model: mV = 1.7 TeV (∼ ICHEP

exclusion)

◮ Optimized tight cut: BDT > 0.45

◮ Modeling degrades significantly past 0.6

◮ Two categories are fit simultaneously

BDT cut

0.2 0.4 0.6 0.8 1

theory

σ /

95% CL

σ

0.02 0.04 0.06 0.08 0.1 0.12

CMSPreliminary

(13 TeV)

1

36.6 pb

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 20 / 28

SLIDE 21

Transfer factors (Z → νν)

Top: loose; bottom: tight Dimuon

U [GeV]

300 400 500 600 700 800 900 1000

µ µ → /Z ν ν → Z

1.5 2 2.5 3 3.5 4 4.5 5

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Muons (13 TeV)

1

36.6 fb

CMS Preliminary U [GeV]

300 400 500 600 700 800 900 1000

µ µ → /Z ν ν → Z

1.5 2 2.5 3 3.5 4 4.5 5 5.5

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Muons (13 TeV)

1

36.6 fb

CMS Preliminary

Dielectron

U [GeV]

300 400 500 600 700 800 900 1000

ee → /Z ν ν → Z

2 3 4 5 6 7

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Electrons (13 TeV)

1

36.6 fb

CMS Preliminary U [GeV]

300 400 500 600 700 800 900 1000

ee → /Z ν ν → Z

2 3 4 5 6 7 8

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Electrons (13 TeV)

1

36.6 fb

CMS Preliminary

Photon

U [GeV]

300 400 500 600 700 800 900 1000

γ / ν ν → Z

0.1 0.15 0.2 0.25 0.3

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Photons (13 TeV)

1

36.6 fb

CMS Preliminary U [GeV]

300 400 500 600 700 800 900 1000

γ / ν ν → Z

0.1 0.15 0.2 0.25 0.3 0.35

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Photons (13 TeV)

1

36.6 fb

CMS Preliminary

W/Z

U [GeV]

300 400 500 600 700 800 900 1000

ν (l) → /W ν ν → Z

1 2 3 4 5 6 7

Transfer Factor (Stat Uncert) Stat + Sys Uncert

(13 TeV)

1

36.6 fb

CMS Preliminary U [GeV]

300 400 500 600 700 800 900 1000

ν (l) → /W ν ν → Z

1 2 3 4 5 6 7

Transfer Factor (Stat Uncert) Stat + Sys Uncert

(13 TeV)

1

36.6 fb

CMS Preliminary

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 21 / 28

SLIDE 22

Transfer factors (t¯ t)

Top: loose; bottom: tight Top muon

U [GeV]

300 400 500 600 700 800 900 1000

ν µ b → /t ν b(l) → t

0.2 0.3 0.4 0.5 0.6 0.7

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Muons (13 TeV)

1

36.6 fb

CMS Preliminary U [GeV]

300 400 500 600 700 800 900 1000

ν µ b → /t ν b(l) → t

0.2 0.3 0.4 0.5 0.6 0.7

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Muons (13 TeV)

1

36.6 fb

CMS Preliminary

Top electron

U [GeV]

300 400 500 600 700 800 900 1000

ν be → /t ν b(l) → t

0.2 0.4 0.6 0.8 1

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Electrons (13 TeV)

1

36.6 fb

CMS Preliminary U [GeV]

300 400 500 600 700 800 900 1000

ν be → /t ν b(l) → t

0.2 0.4 0.6 0.8 1

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Electrons (13 TeV)

1

36.6 fb

CMS Preliminary

W muon

U [GeV]

300 400 500 600 700 800 900 1000

ν µ (b) → /t ν b(l) → t

0.5 1 1.5 2 2.5

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Muons (13 TeV)

1

36.6 fb

CMS Preliminary U [GeV]

300 400 500 600 700 800 900 1000

ν µ (b) → /t ν b(l) → t

0.5 1 1.5 2 2.5 3

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Muons (13 TeV)

1

36.6 fb

CMS Preliminary

W electron

U [GeV]

300 400 500 600 700 800 900 1000

ν (b)e → /t ν b(l) → t

0.5 1 1.5 2 2.5 3 3.5 4

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Electrons (13 TeV)

1

36.6 fb

CMS Preliminary U [GeV]

300 400 500 600 700 800 900 1000

ν (b)e → /t ν b(l) → t

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Electrons (13 TeV)

1

36.6 fb

CMS Preliminary

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 22 / 28

SLIDE 23

Transfer factors (W)

Top: loose; bottom: tight W muon

U [GeV]

300 400 500 600 700 800 900 1000

ν µ → /W ν (l) → W

0.1 0.2 0.3 0.4 0.5

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Muons (13 TeV)

1

36.6 fb

CMS Preliminary U [GeV]

300 400 500 600 700 800 900 1000

ν µ → /W ν (l) → W

0.1 0.2 0.3 0.4 0.5 0.6 0.7

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Muons (13 TeV)

1

36.6 fb

CMS Preliminary

W electron

U [GeV]

300 400 500 600 700 800 900 1000

ν e → /W ν (l) → W

0.2 0.4 0.6 0.8 1

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Electrons (13 TeV)

1

36.6 fb

CMS Preliminary U [GeV]

300 400 500 600 700 800 900 1000

ν e → /W ν (l) → W

0.2 0.4 0.6 0.8 1 1.2

Transfer Factor (Stat Uncert) Stat + Sys Uncert

Electrons (13 TeV)

1

36.6 fb

CMS Preliminary

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 23 / 28

SLIDE 24

Post-fit pulls

PhotonEff SFTopMisTag SFTopTag btag ca15_jec ewk0 ewk1 ewk2 ewk3 ewk4 ewk5 ewk6 ewk_loose0 ewk_loose1 ewk_loose2 ewk_loose3 ewk_loose4 ewk_loose5 ewk_loose6 facscale gjethf lumi met mistag loose_stat_dielectron_1 loose_stat_dielectron_2 loose_stat_dielectron_3 loose_stat_dielectron_4 loose_stat_dielectron_5 loose_stat_dielectron_6 loose_stat_dielectron_7 loose_stat_dimuon_1 loose_stat_dimuon_2 loose_stat_dimuon_3 loose_stat_dimuon_4 loose_stat_dimuon_5 loose_stat_dimuon_6 loose_stat_dimuon_7 loose_stat_photon_1 loose_stat_photon_2 loose_stat_photon_3 loose_stat_photon_4 loose_stat_photon_5 loose_stat_photon_6 loose_stat_photon_7 loose_stat_singleelectrontop_0 loose_stat_singleelectrontop_1 loose_stat_singleelectrontop_2 loose_stat_singleelectrontop_3 loose_stat_singleelectrontop_4 loose_stat_singleelectrontop_5 loose_stat_singleelectrontop_6 loose_stat_singleelectronw_1 loose_stat_singleelectronw_2 loose_stat_singleelectronw_3 loose_stat_singleelectronw_4 loose_stat_singleelectronw_5 loose_stat_singleelectronw_6 loose_stat_singleelectronw_7 loose_stat_singleelectronwtop_0 loose_stat_singleelectronwtop_1 loose_stat_singleelectronwtop_2 loose_stat_singleelectronwtop_3 loose_stat_singleelectronwtop_4 loose_stat_singleelectronwtop_5 loose_stat_singleelectronwtop_6 loose_stat_singlemuontop_0 loose_stat_singlemuontop_1 loose_stat_singlemuontop_2 loose_stat_singlemuontop_3 loose_stat_singlemuontop_4 loose_stat_singlemuontop_5 loose_stat_singlemuontop_6 loose_stat_singlemuonw_1 loose_stat_singlemuonw_2 loose_stat_singlemuonw_3 loose_stat_singlemuonw_4 loose_stat_singlemuonw_5 loose_stat_singlemuonw_6 loose_stat_singlemuonw_7 loose_stat_singlemuonwtop_0 loose_stat_singlemuonwtop_1 loose_stat_singlemuonwtop_2 loose_stat_singlemuonwtop_3 loose_stat_singlemuonwtop_4 loose_stat_singlemuonwtop_5 loose_stat_singlemuonwtop_6 loose_stat_wz_1 loose_stat_wz_2 loose_stat_wz_3 loose_stat_wz_4 loose_stat_wz_5 loose_stat_wz_6 loose_stat_wz_7 stat_dielectron_1 stat_dielectron_2 stat_dielectron_3 stat_dielectron_4 stat_dielectron_5 stat_dielectron_6 stat_dielectron_7 stat_dimuon_1 stat_dimuon_2 stat_dimuon_3 stat_dimuon_4 stat_dimuon_5 stat_dimuon_6 stat_dimuon_7 stat_photon_1 stat_photon_2 stat_photon_3 stat_photon_4 stat_photon_5 stat_photon_6 stat_photon_7 stat_singleelectrontop_0 stat_singleelectrontop_1 stat_singleelectrontop_2 stat_singleelectrontop_3 stat_singleelectrontop_4 stat_singleelectrontop_5 stat_singleelectrontop_6 stat_singleelectronw_1 stat_singleelectronw_2 stat_singleelectronw_3 stat_singleelectronw_4 stat_singleelectronw_5 stat_singleelectronw_6 stat_singleelectronw_7 stat_singleelectronwtop_0 stat_singleelectronwtop_1 stat_singleelectronwtop_2 stat_singleelectronwtop_3 stat_singleelectronwtop_4 stat_singleelectronwtop_5 stat_singleelectronwtop_6 stat_singlemuontop_0 stat_singlemuontop_1 stat_singlemuontop_2 stat_singlemuontop_3 stat_singlemuontop_4 stat_singlemuontop_5 stat_singlemuontop_6 stat_singlemuonw_1 stat_singlemuonw_2 stat_singlemuonw_3 stat_singlemuonw_4 stat_singlemuonw_5 stat_singlemuonw_6 stat_singlemuonw_7 stat_singlemuonwtop_0 stat_singlemuonwtop_1 stat_singlemuonwtop_2 stat_singlemuonwtop_3 stat_singlemuonwtop_4 stat_singlemuonwtop_5 stat_singlemuonwtop_6 stat_wz_1 stat_wz_2 stat_wz_3 stat_wz_4 stat_wz_5 stat_wz_6 stat_wz_7 norm_QCDe norm_QCDgamma norm_QCDmu norm_QCDsig norm_stop norm_tt norm_vv pdf renscale sf_ele sf_mu sjbtag sjmistag track_ele track_mu trig_ele trig_pho veto_tau w_ewk0 w_ewk1 w_ewk2 w_ewk3 w_ewk4 w_ewk5 w_ewk6 w_ewk_loose0 w_ewk_loose1 w_ewk_loose2 w_ewk_loose3 w_ewk_loose4 w_ewk_loose5 w_ewk_loose6 wfacscale wjethf wpdf wrenscale zjethf

θ 3 − 2 − 1 − 1 2 3

Prefit B-only fit S+B fit

◮ Biggest pulls: electrons (SF, reco, trigger)

◮ Explained by observed pre-fit discrepancy

◮ Working on a more readable form of this plot

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 24 / 28

SLIDE 25

Control region fit results

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) VV Single top t t ll → Z

Dimuon CR 0.1<BDT<0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) VV Single top t t ll → Z

Dimuon CR BDT>0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) VV Single top t t ll → Z

Dielectron CR 0.1<BDT<0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) VV Single top t t ll → Z

Dielectron CR BDT>0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) QCD +jets γ

Photon CR 0.1<BDT<0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) QCD +jets γ

Photon CR BDT>0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 25 / 28

SLIDE 26

Control region fit results

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) ll → Z QCD VV Single top ν l → W t t

CR t Single muon t 0.1<BDT<0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) ll → Z QCD VV Single top ν l → W t t

CR t Single muon t BDT>0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) ll → Z QCD VV Single top ν l → W t t

CR t Single electron t 0.1<BDT<0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) ll → Z QCD VV Single top ν l → W t t

CR t Single electron t BDT>0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) QCD VV ll → Z Single top t t ν l → W

Single muon W CR 0.1<BDT<0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

3

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) QCD VV ll → Z Single top t t ν l → W

Single muon W CR BDT>0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) QCD VV ll → Z Single top t t ν l → W

Single electron W CR 0.1<BDT<0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

Events / GeV

3 −

10

2 −

10

1 −

10 1 10

2

10

Data SM backgrounds (pre-fit) SM backgrounds (post-fit) QCD VV ll → Z Single top t t ν l → W

Single electron W CR BDT>0.45

(13 TeV)

1

36.6 fb

CMS Preliminary

Recoil [GeV]

300 400 500 600 700 800 900 1000 Data / Pred.

0.5 1 1.5 2 pre-fit post-fit

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 26 / 28

SLIDE 27

Expected limit [on mV , mχ]

◮ Using a rough grid for now

◮ Interpolation is coming

◮ Expected exclusion

mV ∼ 1.8 TeV

◮ NB: choosing purely vector

couplings

◮ Different from ICHEP (V+A)!

◮ To-do: use the ME reweights to

scan couplings

◮ Can turn on gA and recover

V+A

[GeV]

V

m

200 400 600 800 1000 1200 1400 1600 1800 2000 2200

[GeV]

χ

m

200 400 600 800 1000 1200

2 −

10

1 −

10 1 10

2

10

= 1

V DM

= 0.25, g

V q

g Median Expected 95% CL 1 std. dev. (exp) ± Exp.

(13 TeV)

1

36.3 fb

theory

σ /

95% CL

σ Expected

CMSPreliminary

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 27 / 28

SLIDE 28

Conclusion

To-do list:

◮ Update documentation with results shown in this talk ◮ Rerun analysis with re-MINIAOD and latest corrections ◮ Non-resonant signal model checks

◮ Scan couplings ⇒ set limits on couplings ◮ Mass exclusions for different coupling scenarios (V,A,V+A,V-A) ◮ Check scaling of limits from ICHEP analysis

◮ Add resonant signal interpretation

◮ Model has not changed since previous result

Conclusions:

◮ Modulo a few discrepancies, the control regions are in good shape ◮ Substantial improvements have been made to the sensitivity of the analysis (top-tagging,

background estimation)

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 28 / 28

SLIDE 29

BACKUP

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 29 / 28

SLIDE 30

Leptons and photons

“loose” used for vetos in signal region “tight” used to define control regions Object min pT [GeV] max |η| Other requirements Muons “loose” 10 2.4 Loose ID/Iso “tight” 20 2.4 Tight ID/Iso Electrons* “loose” 10 2.5 Veto ID/Iso “tight” 40 2.4 Tight ID/Iso Taus “loose” 18 2.3 decayModeFinding ∆β Corr3Hits < 5 GeV Photons “loose” 15 2.5 Loose ID/Iso “tight” 175 1.4442 Medium ID/Iso

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 30 / 28

SLIDE 31

Emiss

T

and hadronic recoil

Emiss

T

is main observable for this analysis

◮ Corrections applied to narrow

jets are also applied to Emiss

T

Hadronic recoil U

◮ Defined as

U =

Emiss

T

+ V , where V is:

◮ One lepton for single-lepton CRs ◮ Two leptons for di-lepton CRs ◮ Photon for photon CR

◮ To model Emiss T

shape of SM backgrounds in SR, U shape in CRs is used

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 31 / 28

SLIDE 32

Data and triggers

◮ 36 fb−1 from

Run2016[BCDEFGH]

◮ PromptReco ◮ MET ⇒ signal and muon

control regions

◮ SingleElectron ⇒

electron control regions

◮ SinglePhoton ⇒ photon

control region

Dataset Triggers MET PFMET170_NoiseCleaned PFMET170_JetIdCleaned PFMET170_HBHECleaned PFMETNoMu90_PFMHTNoMu90_IDTight PFMETNoMu100_PFMHTNoMu100_IDTight PFMETNoMu110_PFMHTNoMu110_IDTight PFMETNoMu120_PFMHTNoMu120_IDTight SingleElectron Ele25_eta2p1_WPTight_Gsf Ele27_eta2p1_WPLoose_Gsf Ele27_WPTight_Gsf Ele35_WPLoose_Gsf SinglePhoton Photon165_HE10 Photon175 Photon300_NoHE

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 32 / 28

SLIDE 33

Summary of scale factors and systematics

◮ NLO corrections to V +jets ◮ V +HF uncertainty ◮ Top-tag and top-mistag SFs ◮ b-tag and subjet-b-tag SFs

◮ Using POG recommendations ◮ Uncertainties applied as shape variations

◮ Trigger efficiencies

◮ 2% systematic for each of electron and

photon triggers

◮ Emiss T

uncertainty: 5%

◮ CA15 JEC ◮ Luminosity: 6.2% ◮ Lepton scale factors

◮ ID and iso SFs for electrons and muons

are same as used in monojet analysis

◮ 2% per-leg uncertainty ◮ Separate nuisances for electrons and

muons

◮ Also applying POG-recommended

tracking SFs

◮ NPV-dependent ◮ 1% per-leg uncertainty ◮ Separate nuisances for electrons and

muons

◮ Tau veto: 3% ◮ Photon ID SF: 2%

S. Narayanan (MIT)

Hadronic monotop 24/02/2017 33 / 28

SLIDE 34

Fit likelihood

L(µ µ µZ→νν,µ µ µt¯

t, µ;θ

θ θ) =

i∈bins

Poisson

dℓℓ

i

Bℓℓ

i (θ

θ θ) + µZ→νν

i

Rℓℓ

i (θ

θ θ)

×
i∈bins

Poisson

dγ

i

Bγ

i (θ

θ θ) + µZ→νν

i

Rγ

i (θ

θ θ)

×
i∈bins

Poisson

dbℓ

i

Bbℓ

i (θ

θ θ) + µt¯

t

Rbℓ

i (θ

θ θ)

×
i∈bins

Poisson

dℓ

i

Bℓ

i(θ

θ θ) + fi(θ θ θ)µZ→νν

i

Rℓ

i(θ

θ θ) + µt¯

t

R(b)ℓ

i

(θ θ θ)

×
i∈bins

Poisson

dsignal

i

Bsignal

i

(θ θ θ) + (1 + fi(θ θ θ))µZ→νν

i

+ µt¯

t + µSi(θ

θ θ)

(1)
S. Narayanan (MIT)

Hadronic monotop 24/02/2017 34 / 28