1
00:00:05,100 --> 00:00:06,990
No, let me share my screen.

2
00:00:12,660 --> 00:00:15,780
Can you see my screen? Yeah, it works. Go
ahead.

3
00:00:17,070 --> 00:00:21,900
Okay, so yeah, ehm today I'm going to be
talking about recent ttH measurement with

4
00:00:21,900 --> 00:00:23,460
the Atlas experiment.

5
00:00:24,990 --> 00:00:25,890
And, oops,

6
00:00:27,450 --> 00:00:32,790
okay, we'll start with a brief motivation
of why we're interested with ttH

7
00:00:32,790 --> 00:00:37,890
measurements. As we know all the Higgs property
measurements so far are consistent with

8
00:00:37,890 --> 00:00:43,050
standard model, and one of the information from
measurements are the couplings of

9
00:00:43,050 --> 00:00:48,240
the fermions, in particular the top Yukawa
coupling, which is the goal of this ttH

10
00:00:48,240 --> 00:00:51,540
analysis, and we can directly measure it
via ttH cross section

11
00:00:51,540 --> 00:00:56,670
measurements. Since the top Yukawa coupling
is the largest in standard model as it's

12
00:00:56,670 --> 00:01:00,720
shown on the top right plot, it has a
major impact on the theory, such as the

13
00:01:00,720 --> 00:01:06,090
impact on the effective potential of the Higgs
field and which the variations on

14
00:01:06,090 --> 00:01:09,780
coupling can lead to second minima that
has an impact on the understanding of the

15
00:01:09,780 --> 00:01:16,200
metastability of the universe as
can be seen on the bottom right plot, and

16
00:01:16,200 --> 00:01:20,310
also has an impact on self coupling of the
Higgs which is still to be measured and

17
00:01:20,340 --> 00:01:25,230
a major milestone of the future LHC
experiments. Another interesting point is

18
00:01:25,260 --> 00:01:30,390
CP nature of this coupling, where the
standard model predicts CP even Higgs boson,

19
00:01:30,690 --> 00:01:36,540
but on the other hand, the CP odd
is not excluded yet. And on bottom plot, you

20
00:01:36,540 --> 00:01:40,860
can see the total invariant mass
distribution ending on the mixing angle, where

21
00:01:40,860 --> 00:01:45,870
difference can change the peak of mass
distribution of ttH. And this talk is

22
00:01:45,870 --> 00:01:52,980
actually given by another session in this
in this conference. So, the ttH

23
00:01:52,980 --> 00:01:57,090
measurements at ATLAS experiment utilizes
the broad range of final states

24
00:01:57,090 --> 00:02:02,040
originating from ttbar process and
exploit Higgs decay mode. Therefore the

25
00:02:02,040 --> 00:02:06,720
measurements are categorized depending on
the decay modes of Higgs boson, on bottom

26
00:02:06,720 --> 00:02:12,630
table I'm giving you showing you those
categories. First category is the Higgs to bb

27
00:02:12,630 --> 00:02:16,050
channel. This channel has much high
rate, suffering from the enormous

28
00:02:16,050 --> 00:02:20,970
QCD background and the latest publication
is with 36 inverse femtobarn. And the

29
00:02:20,970 --> 00:02:25,860
second group is called a ttH multilepton
group, where it focuses on WZ boson

30
00:02:25,860 --> 00:02:29,580
and tau decays of Higgs boson. And
the rate is reasonable but the

31
00:02:29,580 --> 00:02:32,550
background estimation is also
challenging due to non prompt backgrounds,

32
00:02:33,180 --> 00:02:38,070
and the latest publications is with 18 inverse
femtobarn. The last category is ZZ

33
00:02:38,070 --> 00:02:41,580
and gamma gamma decays where the
decay rates are quite small, with respect to

34
00:02:41,580 --> 00:02:46,980
others. However, they have a clean signal
without backgrounds, and the latest

35
00:02:46,980 --> 00:02:51,630
publication for both of them is with full run
two data set. And the final observed

36
00:02:51,630 --> 00:02:56,700
finally observation of the ttH
measurement announced by ATLAS in 2018 where

37
00:02:56,700 --> 00:03:00,600
only the gamma gamma channels was using
eighty inverse femtobarns while the

38
00:03:00,600 --> 00:03:02,160
other were using 36 inverse femtobarns.

39
00:03:04,259 --> 00:03:08,789
In next slide I'm starting with Higgs to bb channel
and I shouldn't go so much in details since it was a

40
00:03:08,789 --> 00:03:12,899
it was presented many times in this
conference and and on this channel we are

41
00:03:12,929 --> 00:03:16,919
currently working on the updating the
check the results with the full run two

42
00:03:16,919 --> 00:03:21,869
data, just to give some brief information
this channel is most common on other

43
00:03:21,869 --> 00:03:26,609
channel categories, but it has a
challenging final step as I said, and for

44
00:03:26,609 --> 00:03:30,359
this situation, to handle the situation
the various control regions are defined to

45
00:03:30,359 --> 00:03:34,619
constrain those backgrounds, and analysis
regions are split in two with one

46
00:03:34,619 --> 00:03:41,609
lepton and dilepton, and also a boosted one
lepton channel is exploited, with the signal

47
00:03:41,609 --> 00:03:45,779
and control regions are defined depending
on the number of jets and the b-tagging working point.

48
00:03:46,139 --> 00:03:50,969
An example sketch us is given on the
right bottom and the y x and y axes

49
00:03:50,969 --> 00:03:54,869
represent a combination of different
working points for the b-jets and the

50
00:03:54,869 --> 00:03:59,579
tightest will be the highest purity in
ttH. And as it loosen, the control regions

51
00:03:59,579 --> 00:04:04,109
are defined. And the main challenge is the
modelling of the ttbar jet backgrounds

52
00:04:04,409 --> 00:04:09,449
since it's inferred indirectly
with the tt Higgs to bb final state as it can

53
00:04:09,449 --> 00:04:14,579
be seen from the given Feynman diagram here.
And the control regions are depending on

54
00:04:14,579 --> 00:04:18,629
the flavor of the jets. And furthermore,
numbers of the uncertainties are implemented

55
00:04:18,629 --> 00:04:23,069
to cover the modeling differences between
different schemes and the generators. And

56
00:04:23,069 --> 00:04:26,399
also the analysis make use of
sophisticated multivariate techniques to

57
00:04:26,399 --> 00:04:32,279
extract the signal region in which I'm going
to talk about in the next slide. So, two

58
00:04:32,279 --> 00:04:37,469
steps were employed to discriminate
the signal from background. What's

59
00:04:37,469 --> 00:04:41,099
called intermediate techniques are
focusing on the reconstruction of the tt Higgs

60
00:04:41,099 --> 00:04:47,129
system, such as the reconstruction.. extraction of
boosted decision tree to get the best Higgs

61
00:04:47,129 --> 00:04:52,169
and top candidates in the event. And the likelihood
discriminant considers the kinematic

62
00:04:52,169 --> 00:04:56,219
such as the possible jet particle assignments
and calculates the discriminant for each

63
00:04:56,219 --> 00:05:01,829
event again, and the matrix element method
also provides discrimination built from

64
00:05:01,829 --> 00:05:07,109
first principles. On the bottom plot, left
plot, the proper discrimination on the

65
00:05:07,169 --> 00:05:12,929
reconstruction of the Higgs boson mass is given.
And finally, those outputs from the methods

66
00:05:12,929 --> 00:05:17,639
are combined with an additional
information in the second stage in which is

67
00:05:17,639 --> 00:05:22,139
called the classification BDT which will
will be used to extract the signal

68
00:05:22,139 --> 00:05:26,879
strength and on the bottom right plot one
of the signal region in this channel is

69
00:05:26,879 --> 00:05:30,599
shown, where the classification BDT is
used and obviously, discrimination is

70
00:05:30,599 --> 00:05:37,709
quite successful. Coming to the fit setup
tt Higgs bb uses the profile likelihood fit to extract

71
00:05:38,219 --> 00:05:45,029
mu ttH in 10 control region and nine
signal regions. The classification BDT is

72
00:05:45,029 --> 00:05:49,259
used for signal region, and for the control
regions usually a single bin or kinematic

73
00:05:49,259 --> 00:05:55,049
variable such as some of the pT of the
objects. And the major backgrounds are tt plus

74
00:05:55,049 --> 00:06:00,929
jets, are, are free floating in the fit to
control better those backgrounds. And the summary

75
00:06:00,929 --> 00:06:04,499
of the distributions are given on the
right side of the slides and showing the

76
00:06:04,499 --> 00:06:09,659
comparison of expected and observed yields
and overall very good agreement with data

77
00:06:09,659 --> 00:06:14,699
is observed and the results for each
channel is given on the table bottom left.

78
00:06:14,789 --> 00:06:21,569
And the significance is 1.4 sigma observed
with 1.6 sigma expected. And the major

79
00:06:21,569 --> 00:06:26,459
backgrounds are arising from tt plus jets
modeling and in the from the Monte Carlo

80
00:06:26,459 --> 00:06:34,199
statistics. And obviously the analysis is
dominated by systematic uncertainties. So,

81
00:06:34,199 --> 00:06:39,089
I now I move to the second analysis categories
called ttH multilepton. And this analysis is

82
00:06:39,089 --> 00:06:45,269
focusing on the Higgs to WW, ZZ and tau tau decay modes,
together with ttbar system, requiring

83
00:06:45,719 --> 00:06:50,879
at least one lepton. The baseline selection
of these analysis channels are done

84
00:06:50,879 --> 00:06:55,529
depending on the charge and the flavor of
the leptons. There are six channels are

85
00:06:55,529 --> 00:07:00,929
used, three of them with and without hadronic
taus in the final states, as shown on the

86
00:07:00,929 --> 00:07:05,909
table right, and the most sensitive
channels are two leptons, same sign, zero tau

87
00:07:05,909 --> 00:07:11,129
and the three leptons zero taus, are also further
splitted into control regions to constrain

88
00:07:11,129 --> 00:07:16,439
the different different sources of the
backgrounds. The main backgrounds are

89
00:07:16,439 --> 00:07:20,969
standard model processes such as the ttW,
ttZ, diboson and also the non prompt

90
00:07:20,969 --> 00:07:26,999
backgrounds, mainly coming from ttbar
process, such as leptons arising

91
00:07:26,999 --> 00:07:31,259
from semileptonic b decays or the
conversions or electrons where

92
00:07:31,409 --> 00:07:37,799
one of the electron has charged misassigned
and for to to further reduce those

93
00:07:37,799 --> 00:07:42,749
background contributions there is special
object selections are used on top of

94
00:07:42,749 --> 00:07:49,889
the baseline selections. At the BDT for
instance, BDT isolation is employed to

95
00:07:49,889 --> 00:07:55,529
reduce the leptons from heavy decay
flavors, flavor decays, and a charge misid BDT

96
00:07:55,529 --> 00:08:01,679
also provided by performance group is used,
and variables to reject the material and gamma star

97
00:08:01,709 --> 00:08:03,629
conversions are also benefited.

98
00:08:05,040 --> 00:08:08,760
Just to be a bit more concrete, in this
slide, I'm showing you the region

99
00:08:08,760 --> 00:08:14,100
definition on most sensitive channels in
ttH multilepton. And as well as I will give some

100
00:08:14,100 --> 00:08:19,110
details about non prompt background
estimation. And regions are further divided

101
00:08:19,110 --> 00:08:24,780
into high and low N jet regions. In high
N jet regions the multi dimensional BDT

102
00:08:24,780 --> 00:08:30,420
space is exploit for signal and control
region, and also further split on depending

103
00:08:30,420 --> 00:08:34,740
on the charge and the flavour of the leptons
to constrain different source of

104
00:08:34,770 --> 00:08:40,020
backgrounds again, and it contains again
signal regions in high BDT and control

105
00:08:40,020 --> 00:08:45,420
regions. And the low N jet regions are
more dedicated to control the non prompt

106
00:08:45,420 --> 00:08:50,490
backgrounds and also the ttW standard model
process and there are some kinematic

107
00:08:50,490 --> 00:08:55,290
variables such as HT lepton, N b jet
distribution and also events yield is used.

108
00:08:55,740 --> 00:09:00,840
And on the right plot I'm showing you one
of these regions in this category. There's

109
00:09:00,840 --> 00:09:06,180
sum of the pT of the leptons, here the
orange contribution is non prompt muons and

110
00:09:06,180 --> 00:09:11,760
is populated in the first two bins. It
gives a good control of this background.

111
00:09:12,480 --> 00:09:16,710
And the estimation of light background,
light lepton background, is based on the

112
00:09:16,710 --> 00:09:20,940
Monte Carlo shapes. And the normalizations
are obtained from simultaneous fit to

113
00:09:20,940 --> 00:09:26,790
signal strength. And hum here I'm giving
you these processes. Heavy flavor semileptonic

114
00:09:26,820 --> 00:09:32,280
decay, where an isolated lepton can
be reconstructed within the heavy flavor and

115
00:09:32,700 --> 00:09:36,360
sorry, and dedicated normalization factors
for muons and

116
00:09:36,450 --> 00:09:37,920
electrons are defined.

117
00:09:38,670 --> 00:09:43,230
The second is the electron from material
conversion, where the gamma interact with

118
00:09:43,230 --> 00:09:48,150
detector material and resulting in one is
isolated lepton and the last one is

119
00:09:48,150 --> 00:09:53,190
electron from internal conversions, where
the virtual gamma star decays into two

120
00:09:53,190 --> 00:09:59,550
leptons, again, one being isolated. And it
can occur in at a parton showel, shower or

121
00:09:59,550 --> 00:10:05,790
the matrix element depending on generator. On
slide number nine I want to give more

122
00:10:05,790 --> 00:10:11,730
details about ttW measurements within the
ttH multilepton channel. The standard

123
00:10:11,730 --> 00:10:16,440
model's ttW process is free floated
in the simultaneous fit due to the high

124
00:10:16,440 --> 00:10:21,240
contamination in both control and signal
regions. Ehm three decorrelated normalization

125
00:10:21,240 --> 00:10:26,940
factors are assigned: two lepton same sign low
N jets, high N jet, and three lepton. And

126
00:10:26,940 --> 00:10:30,270
the reason why we divide the normalization
factors is that the different phase space

127
00:10:30,300 --> 00:10:36,360
yields different values and on the right plot
inclusive num ehm N jet distribution for

128
00:10:36,990 --> 00:10:42,900
for two leptons same sign is shown. Here the
dashed lines on the ratio plot in

129
00:10:42,930 --> 00:10:47,550
indicates the pre-fit ratio. And it's
clear that the first two bins needs a

130
00:10:47,550 --> 00:10:53,430
more ttW than the rest to improve the
data Monte Carlo agreement. However, these

131
00:10:53,430 --> 00:10:59,460
results give us a more ttW than the standard
model expected. These values are given on the

132
00:10:59,460 --> 00:11:05,010
table here and those numbers are also
including a missing higher order

133
00:11:05,010 --> 00:11:11,280
corrections to account for NLO ttW
plus jets due to the opening up in the

134
00:11:11,310 --> 00:11:17,910
quark ehm gluon quark initiated process and
also the electroweak corrections where the 

135
00:11:17,910 --> 00:11:24,210
NLO tree term is dominating due to the
tt tW scattering through the Higgs boson and

136
00:11:24,210 --> 00:11:28,920
also were not included into the yellow
report four and as well as the nominal

137
00:11:28,920 --> 00:11:35,250
sample of us and impacts on them can be
non negligible. And finally, this the 

138
00:11:35,250 --> 00:11:41,100
mismodelings are seen observed in a two b plus
plus region. And this difference is added

139
00:11:41,130 --> 00:11:45,990
in the fit model as a shape uncertainty
for the ttW phase space. And the

140
00:11:45,990 --> 00:11:50,610
motivation behind this is purely
experimental. But actually the charges

141
00:11:50,610 --> 00:11:55,620
method is known for a minimum for the ttW
process. And also, there's very

142
00:11:55,620 --> 00:11:56,850
interesting talk in this

143
00:11:58,049 --> 00:11:59,939
session about the ttW modelling.

144
00:12:01,530 --> 00:12:07,200
On the last slide in this channel is for
the fit set up. As usual profile likelihood used

145
00:12:07,200 --> 00:12:12,600
for 17 control regions and eight signal
region as shown in the bottom plot and the

146
00:12:12,600 --> 00:12:17,700
results per channel is given on bottom
right plot. And the overall significance

147
00:12:17,700 --> 00:12:23,760
is 1.8 sigma observed with 3.1 sigma
expected and the major schematics are JES

148
00:12:23,820 --> 00:12:31,620
an tt ll modelling and ttW modeling and overall
good data Monte Carlo agreement is observed. The

149
00:12:31,620 --> 00:12:36,450
next search that I'm going to cover
today is tt Higgs to tt gamma gamma. This

150
00:12:36,450 --> 00:12:41,730
channel is using already full run two data.
And regions are divided into hadronic

151
00:12:41,730 --> 00:12:48,030
where there is one lepton veto there is lepton
veto, and leptonic where there is one or

152
00:12:48,030 --> 00:12:53,340
more leptons in the final state, and of
course with two well isolated photons. Ehm the

153
00:12:53,340 --> 00:12:56,550
background estimation is taken from
sidebands in data where the photon

154
00:12:56,550 --> 00:13:02,250
requirement is reversed and also they
make use of dedicated control samples to

155
00:13:02,250 --> 00:13:08,400
define the function to fit it. And after
certain cuts on BDT to reject the highest

156
00:13:08,400 --> 00:13:13,350
possible background contamination, the bed
regions are further divided depending

157
00:13:13,350 --> 00:13:19,080
on the purity of the signal. And signal
extraction is performed in invariant mass of

158
00:13:19,080 --> 00:13:24,540
diphoton distribution by using double sided
Crystal Ball function. And here on top

159
00:13:24,540 --> 00:13:29,370
right I'm showing you all the regions
included in the fit and bottom post shows

160
00:13:29,370 --> 00:13:34,290
you the fit to the invariant mass of 
diphoton system in all categories, where

161
00:13:34,290 --> 00:13:38,490
the events are weighted depending on the
purity of the category. The significance was found - 

162
00:13:38,490 --> 00:13:45,090
you have two minutes - to be, okay. The 
significance is found to be 4.9 sigmas observed which

163
00:13:45,090 --> 00:13:50,550
means this channel has already almost
reached the observation alone. And this

164
00:13:51,450 --> 00:13:57,600
result as we expect is stat limited. The
last channel I'm going to cover today is

165
00:13:57,600 --> 00:14:01,410
the ttH production where the Higgs decays to
ZZ and subsequently to four l.

166
00:14:01,440 --> 00:14:05,460
And in this relevant paper in fact, the cross
sections are measured for main production

167
00:14:05,460 --> 00:14:11,040
modes in several exclusive regions of the
Higgs boson production phase space, but

168
00:14:11,040 --> 00:14:16,020
today I will just talk about only the ttH
production mode. In the analysis is used also make

169
00:14:16,020 --> 00:14:21,840
use of full run two statistics. And
they focus on the Higgs mass window between

170
00:14:22,680 --> 00:14:27,960
115 and 130 for the signal regions, which is
quite low, then the background estimations

171
00:14:27,960 --> 00:14:32,910
are done from the dedicated sidebands and
the analysis regions are classified depending on

172
00:14:32,910 --> 00:14:38,760
the ttbar decays. The event yield is
used for the ttH, for the signal

173
00:14:38,760 --> 00:14:43,770
extraction in ttH lepton enriched region,
and for the ttH hadronic region

174
00:14:44,370 --> 00:14:49,080
sophisticated neural network algorithms
are exploited which are trained against

175
00:14:49,080 --> 00:14:54,840
the back main backgrounds. And as a result,
two indepedendent outputs are used to

176
00:14:54,840 --> 00:15:00,750
extract in this region, and the distributions
are given on the bottom plot. Then the

177
00:15:00,750 --> 00:15:06,300
extracted signal is 1.7 with big statistical
uncertainty as expected, but three

178
00:15:06,300 --> 00:15:10,560
candidates in total are observed in this
channel which is a good success with such

179
00:15:10,590 --> 00:15:14,850
a smaller channel. And finally, I'm coming
back to a combination of ttH

180
00:15:14,850 --> 00:15:19,890
measurements. There is no combination
is performed yet for by exploiting the

181
00:15:19,890 --> 00:15:24,870
latest results that I've shown today.
However, the observation of ttH measurement

182
00:15:24,870 --> 00:15:31,920
was already performed using very using
partially run two data. And here on the

183
00:15:31,950 --> 00:15:36,240
left plot I'm showing you data and
theory comparison of results with the

184
00:15:36,240 --> 00:15:41,310
first point from run one and the second
point from partially run two data and on

185
00:15:41,310 --> 00:15:47,190
right plot there is also all channel
contribute to that observation which is

186
00:15:47,190 --> 00:15:54,570
shown and 5.8 sigma is
observed. And also the the combined with

187
00:15:54,570 --> 00:16:00,930
this result with full with run one and
and have the 6.3 sigma was observe it

188
00:16:00,930 --> 00:16:02,910
takes six sigma observed

189
00:16:04,559 --> 00:16:09,539
Okay so then in conclusion, I showed today
the latest studies related to tt Higgs

190
00:16:09,569 --> 00:16:14,459
measurements. Atlas has established
observation of ttH production by

191
00:16:14,459 --> 00:16:20,009
combining various Higgs decay channels
and results are all compatible standard

192
00:16:20,009 --> 00:16:25,949
model. And some individual channels have updated
their results with more data. Observation,

193
00:16:25,949 --> 00:16:30,479
for example, observation, almost
observation with ttH production in gamma

194
00:16:30,479 --> 00:16:37,229
gamma channel, and also the three events
are observed with ttH to four l channel.

195
00:16:37,829 --> 00:16:43,439
And for, and for the other channels there's
the challenging background is there always

196
00:16:43,439 --> 00:16:48,839
and requires better understanding of the
background modellings, such as ttW,

197
00:16:49,229 --> 00:16:55,739
tt bb, and currently the improvements are
expected with full run two data. And also the

198
00:16:55,739 --> 00:17:00,779
full run two data will allow us a new
opportunities and studies including

199
00:17:01,139 --> 00:17:05,279
differential measurements and also the
property measures. Thank you.

200
00:17:08,490 --> 00:17:12,360
Thank you very much. Any questions?

201
00:17:17,039 --> 00:17:19,439
Yeah, I see Marco has a question. Go
ahead.

202
00:17:20,490 --> 00:17:26,460
Yeah, thank you for this very nice talk.
So, I think in the context of this session

203
00:17:26,460 --> 00:17:31,500
in particular your slide nine is, is very
interesting. So, maybe could could you

204
00:17:31,500 --> 00:17:36,720
take the chance to add some details on the
justification for this say decorrelation,

205
00:17:37,050 --> 00:17:42,540
whether you have done some studies also on
the say generator level phase space, because,

206
00:17:42,810 --> 00:17:49,530
I mean, naively one could wonder
whether this full decorrelation of these

207
00:17:49,620 --> 00:17:58,950
three reco level say phase spaces isn't too
conservative from our point of view, and

208
00:17:58,950 --> 00:18:04,650
then a second part, I mean, to the same
question could be whether you checked

209
00:18:05,190 --> 00:18:12,180
weather you have, for instance, ttW events
that end up confusing with the ttH ones

210
00:18:12,210 --> 00:18:16,710
even at higher jet multiplicities, I
mean here the last bin you show is

211
00:18:16,710 --> 00:18:21,840
the inclusive six jet bin, well, I would
suspect that even higher jet

212
00:18:21,840 --> 00:18:29,340
multiplicities you could play an important
role as a background for ttH in say high

213
00:18:30,690 --> 00:18:33,570
S over B regions. Thank you.

214
00:18:34,890 --> 00:18:40,080
Okay, let me try to get her all. For the
second part I

215
00:18:40,080 --> 00:18:42,630
will say for the ttH ttW

216
00:18:43,890 --> 00:18:48,270
we don't expect to have such events
because I mean in the end, we have the

217
00:18:48,270 --> 00:18:55,080
correlation matrix in after the fit that
we have. We have such not that higher ttW

218
00:18:55,140 --> 00:19:03,240
and ttH uhm correlations in both
normalization factors. Uhm and for the first

219
00:19:03,540 --> 00:19:06,090
point is I'm not sure I really understood
your

220
00:19:06,090 --> 00:19:08,520
question is, I think it was about

221
00:19:09,900 --> 00:19:14,670
is it too conservative to have
decorrelated normalization factors, I guess?

222
00:19:14,970 --> 00:19:21,870
So that the question is for instance, why
do you decorrelate three lepton

223
00:19:22,110 --> 00:19:27,540
from two lepton low jet that in principle, I
would suspect they're populated by the

224
00:19:27,540 --> 00:19:34,230
same say kind of ttW events in a generator
level phase space apart from from the

225
00:19:34,230 --> 00:19:35,100
decay, right.

226
00:19:35,670 --> 00:19:41,220
Okay. So, there was a, we had a dedicated
study for that, in fact,

227
00:19:43,440 --> 00:19:44,070
ehm, while

228
00:19:45,630 --> 00:19:46,860
for example, for the

229
00:19:46,860 --> 00:19:52,920
control region only fit into a three
lepton, and we have also the same

230
00:19:53,880 --> 00:19:58,980
normalization factor of ttW and a
similar one, where we have only one

231
00:19:58,980 --> 00:20:06,030
normalization factor. This three three
l region is is really lowering down in

232
00:20:06,030 --> 00:20:11,370
the I mean its normalisation factor 
is going really low and then in fact

233
00:20:11,370 --> 00:20:17,550
in the ttH measurements and so we found that
decorrelation of three of them is

234
00:20:17,550 --> 00:20:20,490
basically is more reasonable.

235
00:20:32,700 --> 00:20:35,820
Okay, are there any other questions?

236
00:20:37,410 --> 00:20:39,900
You can raise your hands if yes.

237
00:20:45,390 --> 00:20:52,020
Okay, I don't see anything else. Okay,
thank you very much Merve and we move