Post on 23-Mar-2021
Combination of searches for pairs of Higgs bosonswith the ATLAS detector
Petar BokanUniversity of Göttingen, Uppsala University
on behalf of the ATLAS Collaboration
Double Higgs Production at Colliders Workshop, 4-8 September 2018
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Higgs boson pair production at the LHCo SM Higgs boson pair production (gluon-gluon fusion - ggF):
Production cross-section small:
σSM = 33.41 fb at√s = 13 TeV
(for mH = 125.09 GeV, at NNLO and including NNLL correctionsand NLO finite top-quark mass effects)
Phys. Rev. Lett. 117 (2016) 012001 10.23731/CYRM-2017-002 LHCHXSWGHH
o Potential non-resonant BSM enhancements(in this report: modified trilinear Higgs self-coupling strength)
o Benchmark BSM resonancehypotheses:o Randall-Sundrum gravitonG→ HH (spin=2)
o S → HH (spin=0)
2/19
Higgs boson self-couplingHiggs-fermion Yukawa coupling
Resonant production
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Higgs boson pair production at the LHCo SM Higgs boson pair production (gluon-gluon fusion - ggF):
Production cross-section small:
σSM = 33.41 fb at√s = 13 TeV
(for mH = 125.09 GeV, at NNLO and including NNLL correctionsand NLO finite top-quark mass effects)
Phys. Rev. Lett. 117 (2016) 012001 10.23731/CYRM-2017-002 LHCHXSWGHH
o Potential non-resonant BSM enhancements(in this report: modified trilinear Higgs self-coupling strength)
o Benchmark BSM resonancehypotheses:o Randall-Sundrum gravitonG→ HH (spin=2)
o S → HH (spin=0)
2/19
Higgs boson self-couplingHiggs-fermion Yukawa coupling
Resonant production
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Higgs boson pair production at the LHCo SM Higgs boson pair production (gluon-gluon fusion - ggF):
Production cross-section small:
σSM = 33.41 fb at√s = 13 TeV
(for mH = 125.09 GeV, at NNLO and including NNLL correctionsand NLO finite top-quark mass effects)
Phys. Rev. Lett. 117 (2016) 012001 10.23731/CYRM-2017-002 LHCHXSWGHH
o Potential non-resonant BSM enhancements(in this report: modified trilinear Higgs self-coupling strength)
o Benchmark BSM resonancehypotheses:o Randall-Sundrum gravitonG→ HH (spin=2)
o S → HH (spin=0)2/19
Higgs boson self-couplingHiggs-fermion Yukawa coupling
Resonant production
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Di-Higgs final states
Di-Higgs decay modes and relativebranching fractions:
bb WW ττ ZZ γγ
bb 33%
WW 25% 4.6%
ττ 7.4% 2.5% 0.39%
ZZ 3.1% 1.2% 0.34% 0.076%
γγ 0.26% 0.10% 0.029% 0.013% 0.0005%
Channels considered forthis combination:
HH → bb̄bb̄: the highest BR,large multijet background
HH → bb̄τ+τ−:relatively large BR, cleaner final state
- τlepτhad (BR: 45.8%)- τhadτhad (BR: 41.9%)
HH → bb̄γγ:small BR, clean signal extraction due toa good γγ mass resolution
3/19
10.23731/CYRM-2017-002
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
(Resolved) HH→4bo Background:∼ 95% multijet and ∼ 5% tt̄
o Data-driven estimation of themultijet background→ 2b+ nj (n > 2) events model 4b
o tt̄ normalization from data
[GeV]HHm200 400 600 800 1000 1200 1400
Dat
a / B
kgd
0.5
1
1.5
Eve
nts
/ 100
GeV
1−10
1
10
210
310
410
510
610
710Data
Multijet
tHadronic t
tSemi-leptonic t
Scalar (280 GeV)
100×SM HH
=1)PlM (800 GeV, k/KKG
=2)PlM (1200 GeV, k/KKG
Stat+Syst Uncertainty
ATLAS
Resolved Signal Region, 2016
-1 = 13 TeV, 24.3 fbs
[GeV]lead2jm
60 80 100 120 140 160 180 200
[GeV
]su
bl2j
m
60
80
100
120
140
160
180
200
2E
vent
s / 2
5 G
eV
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16Simulation ATLAS
Resolved, 2016
-1 = 13 TeV, 24.3 fbs
o Integrated luminosity(2015+2016): 27.5 fb−1
o Final discriminant: mHH
4/19
arXiv:1804.06174
Signal RegionControl RegionSideband Region
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
(SM non-resonant) HH→bbττo A Boosted Decision Tree (BDT) classification is applied after preselection
BDT score1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1
Eve
nts
/ Bin
1
10
210
310
410
510
610
710Data NR HH at exp limitTop-quark
fakeshadτ →jet +(bb,bc,cc)ττ →Z
OtherSM HiggsUncertaintyPre-fit background
ATLAS -113 TeV, 36.1 fb
SLT 2 b-tagshadτlepτ
BDT score
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1Dat
a/P
red.
0.81
1.2BDT score
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1
Eve
nts
/ Bin
1−10
1
10
210
310
410
510 Data NR HH at exp limitTop-quark
fakeshadτ →jet +(bb,bc,cc)ττ →Z
OtherSM HiggsUncertaintyPre-fit background
ATLAS -113 TeV, 36.1 fb
LTT 2 b-tagshadτlepτ
BDT score
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1Dat
a/P
red.
0.81
1.2
o Jets→ τhad fakes: data-driven(Fake Factor/Rate methods)
o tt̄ and Z + bb̄ normalization from datao 3 signal regions (τlepτhad split into singlee/µ trigger and e/µ+ τ trigger)
o BDT score used as a final discriminant 5/19
arXiv:1808.00336
τlepτhad
SM SMτhadτhad
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
(SM non-resonant) HH→bbγγ
0
5
10
15
20Ev
ents
/ 2.5
GeV ATLAS√s = 13TeV, 36.1 fb 1
1 b-tag, tight selection
DataBkg-only fit
110 120 130 140 150 160m [GeV]
0
5
Data
Bkg
0
2
4
6
8
Even
ts / 2
.5 G
eV ATLAS√s = 13TeV, 36.1 fb 1
2 b-tag, tight selection
DataBkg-only fit
110 120 130 140 150 160m [GeV]
024
Data
Bkg
[GeV]jjγγm250 300 350 400 450
Fra
ctio
n of
eve
nts
0
0.05
0.1
0.15
0.2
0.25
0.3
2 b-tag, loose selection SimulationATLAS
H = mjjsolid line indicates that mconstraint is applied
= 260 GeVXm = 300 GeVXm
= 350 GeVXm = 400 GeVXm
+jetsγγSM
o Tight selection applied (pT of the jets, mjj)o Data-driven methods used to estimate thecontinuum backgrounddouble two-dimensional sideband method
o Unbinned maximum-likelihood fits to thedata in the 1-tag and 2-tag regions
o Final discriminant: mγγ (non-resonant)6/19
arXiv:1807.04873
1b-tagtight tight
2b-tags
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Combination strategyo The combination is realized by constructing a combined likelihood functionthat takes into account data, models and systematic uncertainties
o The asymptotic approximation is used for statistical analysis of all∗channels, and the combination
o All the signal regions considered are orthogonal, or have negligible overlap.o Instrumental and luminosity uncertainties correlated across the channels.o The acceptance and the background modeling uncertainties are treated asuncorrelated.
o Theoretical uncertainties on the total signal cross-section are not considered.
∗For the individual bbγγ result pseudo-experiments were used - agreementat a level of 10% (5% for the combination) for all results
7/19
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Non-resonant SM HH production
7/19
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
SM HH production, combined result
0 10 20 30 40 50 60 70 80
ggF
SMσ HH) normalized to → (pp ggFσ95% CL upper limit on
Combined
γγb b→HH
-τ+τb b→HH
bbb b→HH 12.9 20.7 18.5
12.6 14.6 11.9
20.4 26.3 25.1
6.7 10.4 9.2
Obs. Exp. Exp. stat.
ObservedExpected
σ 1±Expected σ 2±Expected
ATLAS Preliminary-1 = 13 TeV, 27.5 - 36.1 fbs
HH) = 33.4 fb→ (pp ggFSMσ
Run-1 ATLAS combination obs (exp): 70 (48) Phys. Rev. D 92, 092004 8/19
obs: 0.22 pbexp: 0.35 pb
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Trilinear Higgs self-coupling variations
8/19
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Varied trilinear Higgs self-couplingHH production modified
(using scale factors: κt = gtt̄H/gSMtt̄H and κλ = λHHH/λ
SMHHH)
A(κt, κλ) = κ2tB + κtκλT
A(1, 0) = B A(1, 1) = B + T A(1, 2) = B + 2TExpress |B|2, |T |2 and (BT ∗ + TB∗) in terms of |A(1, 0)|2, |A(1, 1)|2 and |A(1, 2)|2,
which leads to:
|A(κt, κλ)|2 = a(κt, κλ)|A(1, 0)|2 + b(κt, κλ)|A(1, 1)|2 + c(κt, κλ)|A(1, 2)|2
Any (κt, κλ) combination at LO can be obtainedfrom a linear combination of some 3 (κt 6= 0, κλ) samples!
9/19
TB
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Varied trilinear Higgs self-couplingHH production modified
(using scale factors: κt = gtt̄H/gSMtt̄H and κλ = λHHH/λ
SMHHH)
A(κt, κλ) = κ2tB + κtκλT
A(1, 0) = B A(1, 1) = B + T A(1, 2) = B + 2TExpress |B|2, |T |2 and (BT ∗ + TB∗) in terms of |A(1, 0)|2, |A(1, 1)|2 and |A(1, 2)|2,
which leads to:
|A(κt, κλ)|2 = a(κt, κλ)|A(1, 0)|2 + b(κt, κλ)|A(1, 1)|2 + c(κt, κλ)|A(1, 2)|2
Any (κt, κλ) combination at LO can be obtainedfrom a linear combination of some 3 (κt 6= 0, κλ) samples!
9/19
TB
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Varied trilinear Higgs self-couplingHH production modified
(using scale factors: κt = gtt̄H/gSMtt̄H and κλ = λHHH/λ
SMHHH)
A(κt, κλ) = κ2tB + κtκλT
A(1, 0) = B A(1, 1) = B + T A(1, 2) = B + 2TExpress |B|2, |T |2 and (BT ∗ + TB∗) in terms of |A(1, 0)|2, |A(1, 1)|2 and |A(1, 2)|2,
which leads to:
|A(κt, κλ)|2 = a(κt, κλ)|A(1, 0)|2 + b(κt, κλ)|A(1, 1)|2 + c(κt, κλ)|A(1, 2)|2
Any (κt, κλ) combination at LO can be obtainedfrom a linear combination of some 3 (κt 6= 0, κλ) samples!
9/19
TB
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Varied trilinear Higgs self-couplingHH production modified
(using scale factors: κt = gtt̄H/gSMtt̄H and κλ = λHHH/λ
SMHHH)
A(κt, κλ) = κ2tB + κtκλT
A(1, 0) = B A(1, 1) = B + T A(1, 2) = B + 2TExpress |B|2, |T |2 and (BT ∗ + TB∗) in terms of |A(1, 0)|2, |A(1, 1)|2 and |A(1, 2)|2,
which leads to:
|A(κt, κλ)|2 = a(κt, κλ)|A(1, 0)|2 + b(κt, κλ)|A(1, 1)|2 + c(κt, κλ)|A(1, 2)|2
Any (κt, κλ) combination at LO can be obtainedfrom a linear combination of some 3 (κt 6= 0, κλ) samples!
9/19
TB
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Linear combinationo Showing generator level mHH for:κλ = {0, 1, 2, 20}(other parameters fixed to the SM)
o Different bases tested for linearcombination(e.g. κλ = {0, 1, 2} vs κλ = {0, 1, 20})
o Remaining sample used for validation(very good closure at generator level) 300 400 500 600 700 800 900 1000
[GeV]HHm
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
Arb
itrar
y un
its ATLAS Simulation Work In Progress=13 TeVs
=0λκ=1λκ=2λκ=20λκ
200 300 400 500 600 700 800 900 1000
[GeV]HHm
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
a.u. ATLAS Simulation Work In Progress
=13 TeVs
=20λκ
generated sample={0,1,2}λκlinear combination
200 300 400 500 600 700 800 900 1000
[GeV]HHm
0
5
10
15
20
25
30
35
40
456−10×
a.u. ATLAS Simulation Work In Progress
=13 TeVs
=2λκ
generated sample={0,1,20}λκlinear combination
10/19
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Trilinear Higgs self-coupling scan strategymκλ=xHH , for x = {−20,−19, ..., 20}, at generator level, at LO
obtained using the linear combination :
200 300 400 500 600 700 800 900 [GeV]HHm
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
3−10×
a.u. ATLAS Simulation Work In Progress
= 13 TeVs
=-5λκ
200 300 400 500 600 700 800 900 [GeV]HHm
0
2
4
6
8
10
12
14
16
18
6−10×
a.u. ATLAS Simulation Work In Progress
= 13 TeVs
=2λκ
200 300 400 500 600 700 800 900 [GeV]HHm
0
0.05
0.1
0.15
0.2
0.25
3−10×
a.u. ATLAS Simulation Work In Progress
= 13 TeVs
=5λκ
200 300 400 500 600 700 800 900 [GeV]HHm
0
0.2
0.4
0.6
0.8
1
1.2
1.4
3−10×
a.u. ATLAS Simulation Work In Progress
= 13 TeVs
=10λκ
Weights, binned in mHH , obtained as: mκλ=xHH |bin i/m
κλ=1HH |bin i
200 300 400 500 600 700 [GeV]HHm
10
210
310
=1
λκ HH
/m=
-5λκ HH
m
ATLAS Simulation Work In Progress=13 TeVs
=-5λκ=16.57=1λκσ/=-5λκσLO
200 300 400 500 600 700 [GeV]HHm
1−10
1
10
=1
λκ HH
/m=
2λκ HH
m
ATLAS Simulation Work In Progress=13 TeVs
=2λκ=0.47=1λκσ/=2λκσLO
200 300 400 500 600 700 [GeV]HHm
1
10
210
310=1
λκ HH
/m=
5λκ HH
m
ATLAS Simulation Work In Progress=13 TeVs
=5λκ=2.42=1λκσ/=5λκσLO
200 300 400 500 600 700 [GeV]HHm
1
10
210
310
=1
λκ HH
/m=
10λκ HH
m
ATLAS Simulation Work In Progress=13 TeVs
=10λκ=17.46=1λκσ/=10λκσLO
These weights are applied to the fully reconstructed NLO SM sample to obtainany κλ point, assuming that the LO to NLO factorization does not depend on κλ
11/19
1
2
3
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Trilinear Higgs self-coupling scan strategymκλ=xHH , for x = {−20,−19, ..., 20}, at generator level, at LO
obtained using the linear combination :
200 300 400 500 600 700 800 900 [GeV]HHm
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
3−10×
a.u. ATLAS Simulation Work In Progress
= 13 TeVs
=-5λκ
200 300 400 500 600 700 800 900 [GeV]HHm
0
2
4
6
8
10
12
14
16
18
6−10×
a.u. ATLAS Simulation Work In Progress
= 13 TeVs
=2λκ
200 300 400 500 600 700 800 900 [GeV]HHm
0
0.05
0.1
0.15
0.2
0.25
3−10×
a.u. ATLAS Simulation Work In Progress
= 13 TeVs
=5λκ
200 300 400 500 600 700 800 900 [GeV]HHm
0
0.2
0.4
0.6
0.8
1
1.2
1.4
3−10×
a.u. ATLAS Simulation Work In Progress
= 13 TeVs
=10λκ
Weights, binned in mHH , obtained as: mκλ=xHH |bin i/m
κλ=1HH |bin i
200 300 400 500 600 700 [GeV]HHm
10
210
310
=1
λκ HH
/m=
-5λκ HH
m
ATLAS Simulation Work In Progress=13 TeVs
=-5λκ=16.57=1λκσ/=-5λκσLO
200 300 400 500 600 700 [GeV]HHm
1−10
1
10
=1
λκ HH
/m=
2λκ HH
m
ATLAS Simulation Work In Progress=13 TeVs
=2λκ=0.47=1λκσ/=2λκσLO
200 300 400 500 600 700 [GeV]HHm
1
10
210
310=1
λκ HH
/m=
5λκ HH
m
ATLAS Simulation Work In Progress=13 TeVs
=5λκ=2.42=1λκσ/=5λκσLO
200 300 400 500 600 700 [GeV]HHm
1
10
210
310
=1
λκ HH
/m=
10λκ HH
m
ATLAS Simulation Work In Progress=13 TeVs
=10λκ=17.46=1λκσ/=10λκσLO
These weights are applied to the fully reconstructed NLO SM sample to obtainany κλ point, assuming that the LO to NLO factorization does not depend on κλ
11/19
1
2
3
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Trilinear Higgs self-coupling scan strategymκλ=xHH , for x = {−20,−19, ..., 20}, at generator level, at LO
obtained using the linear combination :
200 300 400 500 600 700 800 900 [GeV]HHm
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
3−10×
a.u. ATLAS Simulation Work In Progress
= 13 TeVs
=-5λκ
200 300 400 500 600 700 800 900 [GeV]HHm
0
2
4
6
8
10
12
14
16
18
6−10×
a.u. ATLAS Simulation Work In Progress
= 13 TeVs
=2λκ
200 300 400 500 600 700 800 900 [GeV]HHm
0
0.05
0.1
0.15
0.2
0.25
3−10×
a.u. ATLAS Simulation Work In Progress
= 13 TeVs
=5λκ
200 300 400 500 600 700 800 900 [GeV]HHm
0
0.2
0.4
0.6
0.8
1
1.2
1.4
3−10×
a.u. ATLAS Simulation Work In Progress
= 13 TeVs
=10λκ
Weights, binned in mHH , obtained as: mκλ=xHH |bin i/m
κλ=1HH |bin i
200 300 400 500 600 700 [GeV]HHm
10
210
310
=1
λκ HH
/m=
-5λκ HH
m
ATLAS Simulation Work In Progress=13 TeVs
=-5λκ=16.57=1λκσ/=-5λκσLO
200 300 400 500 600 700 [GeV]HHm
1−10
1
10
=1
λκ HH
/m=
2λκ HH
m
ATLAS Simulation Work In Progress=13 TeVs
=2λκ=0.47=1λκσ/=2λκσLO
200 300 400 500 600 700 [GeV]HHm
1
10
210
310=1
λκ HH
/m=
5λκ HH
m
ATLAS Simulation Work In Progress=13 TeVs
=5λκ=2.42=1λκσ/=5λκσLO
200 300 400 500 600 700 [GeV]HHm
1
10
210
310
=1
λκ HH
/m=
10λκ HH
m
ATLAS Simulation Work In Progress=13 TeVs
=10λκ=17.46=1λκσ/=10λκσLO
These weights are applied to the fully reconstructed NLO SM sample to obtainany κλ point, assuming that the LO to NLO factorization does not depend on κλ
11/19
1
2
3
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Differences compared to the SM HH searcho Acceptance changes significantly as a function of κλ:
20− 15− 10− 5− 0 5 10 15 20
λκ
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Effi
cien
cy [%
]×
Acc
epta
nce
2015
2016
ATLAS Simulation Preliminarybbb b→=13 TeV, HHs
20− 15− 10− 5− 0 5 10 15 20
λκ
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
Effi
cien
cy [%
]×
Acc
epta
nce
SLThadτlepτ
hadτhadτ
ATLAS Simulation Preliminary-τ+τb b→=13 TeV, HHs
20− 15− 10− 5− 0 5 10 15 20
λκ
5
6
7
8
9
10
11
Effi
cien
cy [%
]×
Acc
epta
nce
1 b-tag
2 b-tags
ATLAS Simulation Preliminaryγγb b→=13 TeV, HHs
variations of the mHH spectrum with κλ:
300 400 500 600 700 800
[GeV]HHm
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
Arb
itrar
y un
its
ATLAS Simulation Preliminary
= 13 TeVs = 0λκ = 2λκ = 5λκ
300 400 500 600 700 800
[GeV]HHm
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
Arb
itrar
y un
its
ATLAS Simulation Preliminary
= 13 TeVs5− = λκ
= 1λκ= 10λκ
o Looser bbγγ selection (softer pT for large κλ)o bbττ : a dedicated BDT, trained on κλ = 20
signal is used since it performs good for all κλpoints.
o The shape of bbγγ discriminant (mγγ)remains independent of κλ, while an additionalloss in sensitivity is observed for bbττ and 4banalyses for large |κλ|. 1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1
BDT Score Cut
0
0.2
0.4
0.6
0.8
1
1.2
Sig
nal F
ract
ion
=-20λκ=-1λκ=1λκ=2λκ=3λκ=4λκ=5λκ=10λκ=20λκ
ATLAS Simulation Work In Progress
hadτhadτ bb→HH
12/19
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Differences compared to the SM HH searcho Acceptance changes significantly as a function of κλ:
20− 15− 10− 5− 0 5 10 15 20
λκ
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Effi
cien
cy [%
]×
Acc
epta
nce
2015
2016
ATLAS Simulation Preliminarybbb b→=13 TeV, HHs
20− 15− 10− 5− 0 5 10 15 20
λκ
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
Effi
cien
cy [%
]×
Acc
epta
nce
SLThadτlepτ
hadτhadτ
ATLAS Simulation Preliminary-τ+τb b→=13 TeV, HHs
20− 15− 10− 5− 0 5 10 15 20
λκ
5
6
7
8
9
10
11
Effi
cien
cy [%
]×
Acc
epta
nce
1 b-tag
2 b-tags
ATLAS Simulation Preliminaryγγb b→=13 TeV, HHs
variations of the mHH spectrum with κλ:
300 400 500 600 700 800
[GeV]HHm
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
Arb
itrar
y un
its
ATLAS Simulation Preliminary
= 13 TeVs = 0λκ = 2λκ = 5λκ
300 400 500 600 700 800
[GeV]HHm
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
Arb
itrar
y un
its
ATLAS Simulation Preliminary
= 13 TeVs5− = λκ
= 1λκ= 10λκ
o Looser bbγγ selection (softer pT for large κλ)o bbττ : a dedicated BDT, trained on κλ = 20
signal is used since it performs good for all κλpoints.
o The shape of bbγγ discriminant (mγγ)remains independent of κλ, while an additionalloss in sensitivity is observed for bbττ and 4banalyses for large |κλ|. 1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1
BDT Score Cut
0
0.2
0.4
0.6
0.8
1
1.2
Sig
nal F
ract
ion
=-20λκ=-1λκ=1λκ=2λκ=3λκ=4λκ=5λκ=10λκ=20λκ
ATLAS Simulation Work In Progress
hadτhadτ bb→HH
12/19
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Differences compared to the SM HH search
[GeV]HHReconstructed m
200 400 600 800 1000 1200 1400
Eve
nts
/ 100
GeV
1
10
210
310
410
510
610
710Data
Multijet
tHadronic t
tSemi-leptonic t
Stat+Syst Uncertainty
=-5 (x 95)λκNR HH
=1 (x 1291)λκNR HH
=10 (x 98)λκNR HH
PreliminaryATLAS -1 = 13 TeV, 27.5 fbs
bbb b→HH Resolved Signal Region, 2015+2016
[GeV]HHReconstructed m200 400 600 800 1000 1200 1400
S /
B
0
0.01
0.02 xBRσNR HH normalized to
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1BDT
1
10
210
310
410
510
610
710
Eve
nts
/ Bin Data
=-5λκNR HH =1λκNR HH =10λκNR HH
Top-quark fakeshadτ →jet
+(bb,bc,cc)ττ →Z OtherSM HiggsUncertaintyPre-fit background
ATLAS Preliminary -1 = 13 TeV, 36.1 fbs
-τ+τb b→HH SLT 2 b-tagshadτlepτ
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1
BDT
0.81
1.2
Dat
a/P
red.
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1BDT
1
10
210
310
410
510
610
710
Eve
nts
/ Bin Data
=-5λκNR HH =1λκNR HH =10λκNR HH
Top-quark fakes (Multi-jets)hadτ →jet
+(bb,bc,cc)ττ →Z )t fakes (thadτ →jet
OtherSM HiggsUncertaintyPre-fit background
ATLAS Preliminary -1 = 13 TeV, 36.1 fbs
-τ+τb b→HH 2 b-tagshadτhadτ
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1
BDT
0.81
1.2
Dat
a/P
red.
[GeV]HHReconstructed m
200 400 600 800 1000 1200 1400
Eve
nts
/ 100
GeV
1
10
210
310
410
510
610
710Data
Multijet
tHadronic t
tSemi-leptonic t
Stat+Syst Uncertainty
=0 (x 663)λκNR HH
=2 (x 2519)λκNR HH
=5 (x 712)λκNR HH
PreliminaryATLAS -1 = 13 TeV, 27.5 fbs
bbb b→HH Resolved Signal Region, 2015+2016
[GeV]HHReconstructed m200 400 600 800 1000 1200 1400
S /
B
00.0020.0040.006 xBRσNR HH normalized to
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1BDT
1
10
210
310
410
510
610
710
Eve
nts
/ Bin Data
=0λκNR HH =2λκNR HH =5λκNR HH
Top-quark fakeshadτ →jet
+(bb,bc,cc)ττ →Z OtherSM HiggsUncertaintyPre-fit background
ATLAS Preliminary -1 = 13 TeV, 36.1 fbs
-τ+τb b→HH SLT 2 b-tagshadτlepτ
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1
BDT
0.81
1.2
Dat
a/P
red.
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1BDT
1
10
210
310
410
510
610
710
Eve
nts
/ Bin Data
=0λκNR HH =2λκNR HH =5λκNR HH
Top-quark fakes (Multi-jets)hadτ →jet
+(bb,bc,cc)ττ →Z )t fakes (thadτ →jet
OtherSM HiggsUncertaintyPre-fit background
ATLAS Preliminary -1 = 13 TeV, 36.1 fbs
-τ+τb b→HH 2 b-tagshadτhadτ
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1
BDT
0.81
1.2
Dat
a/P
red.
13/19
κλ
={−
5,1,
10}
κλ
={0,2,5}
HH → bb̄bb̄ HH → bb̄τlepτhad HH → bb̄τhadτhad
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Limits on the cross-section as a function of κλ
20− 15− 10− 5− 0 5 10 15 20
SMλ /
HHHλ = λκ
2−10
1−10
1
10
210
HH
) [p
b]→
(pp
gg
Fσ
95%
CL
uppe
r lim
it on
SM
(exp.)bbbb (obs.)bbbb (exp.)-τ+τbb (obs.)-τ+τbb
(exp.)γγbb (obs.)γγbb
Combined (exp.)Combined (obs.)
(Combined)σ1±Expected (Combined)σ2±Expected
NLO theory pred.
ATLAS Preliminary
-127.5 - 36.1 fb = 13 TeV,s
The scale factor κλ is observed (expected) to be constrained in the range:−5.0 < κλ < 12.1 (−5.8 < κλ < 12.0)
14/19
4bbbττbbγγcombination
dashed:expectedsolid:observed
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Full systematic uncertainty vs data stat-only
20− 15− 10− 5− 0 5 10 15 20
SMλ /
HHHλ = λκ
2−10
1−10
1
10
210
HH
) [p
b]→
(pp
gg
Fσ
95%
CL
uppe
r lim
it on
SM
(exp.)bbbb
(exp.)-τ+τbb
(exp.)γγbb
Combined (exp.)Combined (exp.) full syst.
(Combined)σ1±Expected (Combined)σ2±Expected
NLO theory pred.
ATLAS Preliminary
-127.5 - 36.1 fb = 13 TeV,s
Stat. only
Stat. only limits for the individual channels and the combination15/19
4bbbττbbγγcombination
solid:expectedfull syst.dashed:expectedstat-only
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Resonant HH production(combination in the mass range: 260-1000 GeV)
Differences compared to the SM HH search:
bbγγ:looser selection below 500 GeV
final discriminant: mγγjj
bb̄ττ :dedicated BDTs
bb̄bb̄:boosted analysis for signal masses > 800 GeV
(combined with the resolved)Looking for two Higgs candidates, each composed of a single
large-R (1.0) jet with at least one b-tagged track jet associated to it15/19
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Scalar resonance
300 400 500 600 700 800 900 1000 [GeV]Sm
3−10
2−10
1−10
1
10
210
310
HH
) [p
b]→
S
→(p
p σ
95%
CL
uppe
r lim
it on
SM
(exp.)bbbb (obs.)bbbb (exp.)-τ+τbb (obs.)-τ+τbb
(exp.)γγbb (obs.)γγbb
Combined (exp.)Combined (obs.)
(Combined)σ1±Expected (Combined)σ2±Expected
= 2)βhMSSM (tan
ATLAS Preliminary-1 = 13 TeV, 27.5 - 36.1 fbs spin-0
hMSSM, narrow width CP -even Higgs boson (tanβ = 2)∗: mS < 462 GeV at 95% CL∗tanβ = 2: ratio of the vacuum expectation values of the two Higgs doublets
16/19
4bbbττbbγγcombination
dashed:expectedsolid:observed
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Randall-Sundrum graviton model
300 400 500 600 700 800 900 1000 [GeV]
KK
*G
m
2−10
1−10
1
10
210
HH
) [p
b]→
KK
* G
→(p
p σ
95%
CL
uppe
r lim
it on
SM
(exp.)bbbb (obs.)bbbb (exp.)-τ+τbb (obs.)-τ+τbb
Combined (exp.)Combined (obs.)
(Combined)σ1±Expected (Combined)σ2±Expected = 1.0PlMBulk RS, k/
ATLAS Preliminary-1 = 13 TeV, 27.5 - 36.1 fbs = 1.0PlMspin-2, k/
(k/M̄Pl = 1)∗ 307 < mG < 1362 GeV∗∗ at 95% CL∗k: curvature of the warped extra dimension, M̄Pl: the effective four-dimensional Planck scale
∗∗the upper limit on the mass comes from 4b only 17/19
4bbbττcombination
dashed:expectedsolid:observed
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Randall-Sundrum graviton model
300 400 500 600 700 800 900 1000 [GeV]
KK
*G
m
3−10
2−10
1−10
1
10
210
HH
) [p
b]→
KK
* G
→(p
p σ
95%
CL
uppe
r lim
it on
SM
(exp.)bbbb (obs.)bbbb (exp.)-τ+τbb (obs.)-τ+τbb
Combined (exp.)Combined (obs.)
(Combined)σ1±Expected (Combined)σ2±Expected = 2.0PlMBulk RS, k/
ATLAS Preliminary-1 = 13 TeV, 27.5 - 36.1 fbs = 2.0PlMspin-2, k/
(k/M̄Pl = 2)∗ mG < 1744 GeV∗∗ at 95% CL∗k: curvature of the warped extra dimension, M̄Pl: the effective four-dimensional Planck scale
∗∗the upper limit on the mass comes from 4b only 18/19
4bbbττcombination
dashed:expectedsolid:observed
Introduction Channels Combination SM HH H self-coupling Resonant Conclusion
Conclusiono A statistical combination of searches for non-resonant and resonantproduction of Higgs boson pairs for the most sensitive channels is presented.
o Using up to 36.1 fb−1 of pp collision data recorded with the ATLASdetector.
o The observed (expected) 95% CL exclusion upper limit on the SM Higgsboson pair production is set to 6.7 (10.4) times the SM prediction.
o The observed (expected) exclusion limit as a function of the Higgsself-coupling scale factor, κλ, allows to constrain values in the range:−5.0 < κλ < 12.1 (−5.8 < κλ < 12.0) at 95% CL.
o The exclusion limits are set on the production cross-section of heavy spin-0and spin-2 resonances decaying into a pair of Higgs bosons, in the massrange 260-1000 GeV.
o No significant data excess is found after the combination.
Thank you for your attention!19/19
backup slides
19/19
Allowed intervals for κλ
Search channel Allowed κλ interval at 95% CLobs. exp. exp. stat.
HH → bb̄bb̄ −10.9 – 20.1 −11.6 – 18.7 −9.9 – 16.4
HH → bb̄τ+τ− −7.3 – 15.7 −8.8 – 16.7 −7.8 – 15.4HH → bb̄γγ −8.1 – 13.2 −8.2 – 13.2 −7.7 – 12.7Combination −5.0 – 12.1 −5.8 – 12.0 −5.2 – 11.4
20/19
Randall-Sundrum graviton model, 4b
) [TeV]KK
m(G
0.3 0.4 0.5 1 2 3
) [fb
]bbbb
→H
H→
KK
G→
(pp
σ
1
10
210
310
410Observed 95% CL limit
Expected 95% CL limit
σ 1±
σ 2±
= 1PlMBulk RS, k/
ATLAS-1=13 TeV, 27.5-36.1 fbs
bbb b→ HH → KKG
) [TeV]KK
m(G
0.3 0.4 0.5 1 2 3)
[fb]
bbbb→
HH
→K
KG
→(p
pσ
1
10
210
310
410Observed 95% CL limit
Expected 95% CL limit
σ 1±
σ 2±
= 2PlMBulk RS, k/
ATLAS-1=13 TeV, 27.5-36.1 fbs
bbb b→ HH → KKG
21/19