B. BDT input variables 155
B.2. Data-MC Agreement
2l os,
t t t
t tt+light tt+c tt+b
(a)
6 7 8 9 10
Nj
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Norm. Entries
3b 6j, ≥ 2l os, ≥
tt
tt tt+light tt+c tt+b
(b)
1 1.5 2 2.5 3 3.5
Avg
Rjj
∆ 0
0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Norm. Entries
≥ 3b 6j,
≥ 2l os,
t t t
t tt+light tt+c tt+b
(c)
16 18 20 22 24 26 28 30
(PC b-tagging score)i i=1
∑
60 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Norm. Entries
≥ 3b 6j,
≥ 2l os,
t t t
t tt+light tt+c tt+b
(d)
Figure B.1.: Separation of highest ranked BDT input variables in the OS channel.
B.2. Data-MC Agreement
Data MC agreement for BDT input variables after applying the additional SFs of the MC-based method.
B.2. Data-MC Agreement
16 18 20 22 24 26 28 30
Sum of mv2 pseudo-continuous btag score 0.5
139.0fb-1 LJETS,8j,3b Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
1 1.5 2 2.5 3 3.5
139.0fb-1 LJETS,8j,3b Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
0 100 200 300 400 500 600 700 800 900 1000
139.0fb-1 LJETS,8j,3b Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
(a)
16 18 20 22 24 26 28 30
Sum of mv2 pseudo-continuous btag score 0.5
139.0fb-1 LJETS,9j,3b Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
1 1.5 2 2.5 3 3.5
139.0fb-1 LJETS,9j,3b Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others
Uncertainty
*: normalised to total Bkg.
0 100 200 300 400 500 600 700 800 900 1000
139.0fb-1 LJETS,9j,3b Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
(b)
16 18 20 22 24 26 28 30
Sum of mv2 pseudo-continuous btag score 0.5
139.0fb-1 10j,3b
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
1 1.5 2 2.5 3 3.5
139.0fb-1 10j,3b
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
0 100 200 300 400 500 600 700 800 900 1000
139.0fb-1 10j,3b
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
(c)
Figure B.2.: The agreement between data and MC of the most important BDT-input variables in the 1L channel in region with exactly 3 b-jets in the (a) 8-jet, (b) 9-jet and (c) at least 10 jet regions. The uncertainty band includes statistical uncertainties.
16 18 20 22 24 26 28 30 Sum of mv2 pseudo-continuous btag score 0.5
139.0fb-1 OS2L,6j,3b Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
1 1.5 2 2.5 3 3.5
139.0fb-1 OS2L,6j,3b Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others
Uncertainty
*: normalised to total Bkg.
0 100 200 300 400 500 600 700 800 900 1000
139.0fb-1 OS2L,6j,3b Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
(a)
16 18 20 22 24 26 28 30
Sum of mv2 pseudo-continuous btag score 0.5
139.0fb-1 OS2L,7j,3b Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
1 1.5 2 2.5 3 3.5
139.0fb-1 OS2L,7j,3b Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
0 100 200 300 400 500 600 700 800 900 1000
139.0fb-1 OS2L,7j,3b Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
(b)
16 18 20 22 24 26 28 30
Sum of mv2 pseudo-continuous btag score 0.5
139.0fb-1 8j,3b
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
1 1.5 2 2.5 3 3.5
139.0fb-1 8j,3b
tt tt+light
1c
tt Single top
V+jets others
Uncertainty
*: normalised to total Bkg.
0 100 200 300 400 500 600 700 800 900 1000
139.0fb-1 8j,3b
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
(c)
Figure B.3.: The agreement between data and MC of the most important BDT-input variables in the OS channel in region with exactly 3 b-jets in the (a) 6-jet, (b) 7-jet and (c) at least 8 jet regions. The uncertainty band includes statistical uncertainties.
B.2. Data-MC Agreement
16 18 20 22 24 26 28 30
Sum of mv2 pseudo-continuous btag score 0.5
139.0fb-1 4b OS2L,6j,≥ Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
1 1.5 2 2.5 3 3.5
139.0fb-1 4b OS2L,6j,≥ Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others
Uncertainty
*: normalised to total Bkg.
0 100 200 300 400 500 600 700 800 900 1000
139.0fb-1 4b OS2L,6j,≥ Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
(a)
16 18 20 22 24 26 28 30
Sum of mv2 pseudo-continuous btag score 0.5 Events 90
ATLAS Internal = 13 TeV s 139.0fb-1
≥4b
t Single top
V+jets others Uncertainty
*: normalised to total Bkg.
1 1.5 2 2.5 3 3.5
139.0fb-1 4b OS2L,7j,≥ Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others
Uncertainty
*: normalised to total Bkg.
0 100 200 300 400 500 600 700 800 900 1000
139.0fb-1 4b OS2L,7j,≥ Pre-Fit
Data tttt
* tt
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
(b)
16 18 20 22 24 26 28 30
Sum of mv2 pseudo-continuous btag score 0.5
139.0fb-1 4b
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
1 1.5 2 2.5 3 3.5
139.0fb-1 4b
tt tt+light
1c
tt Single top
V+jets others
Uncertainty
*: normalised to total Bkg.
0 100 200 300 400 500 600 700 800 900 1000
139.0fb-1 4b
tt tt+light
1c
tt Single top
V+jets others Uncertainty
*: normalised to total Bkg.
(c)
Figure B.4.: The agreement between data and MC of the most important BDT-input variables in the OS channel in region with at least 4 b-jets in the (a) 6-jet, (b) 7-jet and (c) at least 8 jet regions. The uncertainty band includes statistical uncertainties.
APPENDIX C
Fit to Control Regions with ttTRF
This chapter show additional plots from the fit to data in blinded regions using the ttTRF method. The fit-result is used to generate a pseudo-data set in the full phase-space.
Figure C.1.: Pulls and constraints of the fit to data in blinded regions with the ttTRF method.
0 200 400 600 800 1000 1200 1400 1600 1800 2000 +jets (TRF tt
(a) 8j3bH Prefit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt
(b) 8j3bH Postfit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt
(c) 8j3bL Prefit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt
(d) 8j3bL Postfit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt
(e) 9j3bH Prefit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt
(f) 9j3bH Postfit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt tt+X
(g) 9j3bL Prefit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt tt+X
(h) 9j3bL Postfit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] LJETS,8j,≥ Pre-Fit
Data SM4top tt) +jets (TRF tt tt+X
(i) 8jge4b Prefit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] LJETS,8j,≥ Post-Fit
Data SM4top tt) +jets (TRF tt
(j) 8jge4b Postfit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 LJETS,9j,≥ Pre-Fit
Data SM4top tt) +jets (TRF tt
(k) 9jge4b Prefit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 LJETS,9j,≥ Post-Fit
Data SM4top tt) +jets (TRF tt
(l) 9jge4b Postfit
Figure C.2.: Pre- and post-fit comparison of MC to data in the 1L channel of the fit 163
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5
Events Internal
= 13 TeV, 139 fb-1 +jets (TRF tt
(a) ge10j3bH Prefit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 +jets (TRF tt
(b) ge10j3bH Postfit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 +jets (TRF tt
(c) ge10j3bL Prefit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 +jets (TRF tt tt+X
(d) ge10j3bL Postfit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 +jets (TRF tt tt+X
(e) ge10jge4b Prefit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 +jets (TRF tt tt+X
(f) ge10jge4b Postfit Figure C.3.: Pre- and post-fit comparison of MC to data in the 1L channel of the fit
(1L+OS) to data in blinded regions with the ttTRF method. The H (L) stands for 3b≥4b@85%(3b<4b@85%). The uncertainty band includes statistical and systematical uncertainties.
0 200 400 600 800 1000 1200 1400 1600 1800 2000 +jets (TRF tt
(a) 6j3bH Prefit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt
(b) 6j3bH Postfit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt
(c) 6j3bL Prefit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt
(d) 6j3bL Postfit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt
(e) 7j3bH Prefit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt
(f) 7j3bH Postfit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt tt+X
(g) 7j3bL Prefit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] +jets (TRF tt tt+X
(h) 7j3bL Postfit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] OS2L,6j,≥ Pre-Fit
Data SM4top tt) +jets (TRF tt tt+X
(i) 6jge4b Prefit
0 200 400 600 800 1000 1200 1400 1600 1800 2000 [GeV] OS2L,6j,≥ Post-Fit
Data SM4top tt) +jets (TRF tt
(j) 6jge4b Postfit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 OS2L,7j,≥ Pre-Fit
Data SM4top tt) +jets (TRF tt
(k) 7jge4b Prefit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 OS2L,7j,≥ Post-Fit
Data SM4top tt) +jets (TRF tt
(l) 7jge4b Postfit
Figure C.4.: Pre- and post-fit comparison of MC to data in the 1L channel of the fit 165
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 +jets (TRF tt
(a) ge8j3bH Prefit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 +jets (TRF tt
(b) ge8j3bH Postfit
−1−0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 Events 80
ATLAS Internal = 13 TeV, 139 fb-1 +jets (TRF t
(c) ge8j3bL Prefit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 +jets (TRF tt tt+X
(d) ge8j3bL Postfit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 +jets (TRF tt tt+X
(e) ge8jge4b Prefit
1
− −0.8−0.6−0.4−0.2 0 0.2 0.4 0.6 0.8 1 BDT Score 0.5 +jets (TRF tt tt+X
(f) ge10jge4b Postfit Figure C.5.: Pre- and post-fit comparison of MC to data in the 1L channel of the fit
(1L+OS) to data in blinded regions with the ttTRF method. The H (L) stands for 3b≥4b@85%(3b<4b@85%). The uncertainty band includes statistical and systematical uncertainties.
APPENDIX D
Additional Fit Results
This Appendix includes additional plots for the fits to pseudo data in Section D.1 and for the final results in Section D.2.