Fat jets Tilman Plehn Physics case Higgs tagging Top tagging Fat jets for t ¯ tH production Tilman Plehn Heidelberg University Pheno, 5/2010
Fat jets Physics case Tilman Plehn Long death of t ¯ tH , H → b ¯ b [Michael’s talk — Cammin & Schumacher, CMS-TDR and Atlas-CSC worse] Physics case Fraction of events 0.4 – trigger: t → bW + → b ℓ + ν Higgs tagging 0.35 t → ¯ bW − → ¯ reconstruction and rate: ¯ Top tagging bjj 0.3 – continuum background t ¯ tb ¯ b , t ¯ tjj 0.25 [know at NLO] 0.2 – no chance: 0.15 1– combinatorics: m bb from pp → 4 b tag 2 j ℓν 0.1 0.05 0 0 20 40 60 80 100 120 140 160 180 200 Number of choices
Fat jets Physics case Tilman Plehn Long death of t ¯ tH , H → b ¯ b [Michael’s talk — Cammin & Schumacher, CMS-TDR and Atlas-CSC worse] Physics case Fraction of events 0.4 – trigger: t → bW + → b ℓ + ν Higgs tagging 0.35 t → ¯ bW − → ¯ reconstruction and rate: ¯ Top tagging bjj 0.3 – continuum background t ¯ tb ¯ b , t ¯ tjj 0.25 [know at NLO] 0.2 – not a chance: 0.15 1– combinatorics: m H in pp → 4 b tag 2 j ℓν 0.1 2– kinematics: peak-on-peak 0.05 0 0 20 40 60 80 100 120 140 160 180 200 Number of choices 100 100 Events / 10 GeV Events / 10 GeV ttH(120) ttjj 80 80 ttbb (QCD) ttbb (EW) 60 60 40 40 20 20 0 0 0 0 50 50 100 100 150 150 200 200 250 250 300 300 350 350 400 400 m m (GeV) (GeV) bb bb
Fat jets Physics case Tilman Plehn Long death of t ¯ tH , H → b ¯ b [Michael’s talk — Cammin & Schumacher, CMS-TDR and Atlas-CSC worse] Physics case – trigger: t → bW + → b ℓ + ν Fraction of events 0.4 Higgs tagging t → ¯ bW − → ¯ 0.35 reconstruction and rate: ¯ bjj Top tagging 0.3 – continuum background t ¯ tb ¯ b , t ¯ tjj [know at NLO] 0.25 0.2 – not a chance: 0.15 1– combinatorics: m H in pp → 4 b tag 2 j ℓν 0.1 2– kinematics: peak-on-peak 0.05 √ 3– systematics: S / B ∼ 1 / 9 [ S / B irrelevant] 0 0 20 40 60 80 100 120 140 160 180 200 Number of choices 100 100 Events / 10 GeV Events / 10 GeV ttH(120) ttjj 80 80 ttbb (QCD) ttbb (EW) 60 60 40 40 20 20 0 0 0 0 50 50 100 100 150 150 200 200 250 250 300 300 350 350 400 400 m m (GeV) (GeV) bb bb B TDR reconstruction method 8 Likelihood reconstruction S / 7 LH high lumi LH realistic (low lumi) 6 5 4 3 2 1 0 100 100 110 110 120 120 130 130 140 140 m m (GeV) (GeV) H H
Fat jets Higgs tagging Tilman Plehn New strategy for H → bb [Butterworth, Davison, Rubin, Salam; Adam’s and Michael’s talks] Physics case Higgs tagging – desperately needed for light Higgs [2/3 of all Higgses; inclusive CMS S / B ∼ 1 / 80] Top tagging – S: large m bb , boost-dependent R bb B: large m bb only for large R bb S/B: ask for large m bb and small R bb boosted Higgs R bb ∼ 2 m H / p T ∼ 0 . 8 -1 10 1/ σ tot d σ /dp T -2 10 ttH: p T,t -3 10 ttH: p T,H Wjj: p T,j WH: p T,H -4 10 0 100 200 300 400 500 600 700 p T [ GeV ]
Fat jets Higgs tagging Tilman Plehn New strategy for H → bb [Butterworth, Davison, Rubin, Salam; Adam’s and Michael’s talks] Physics case Higgs tagging – desperately needed for light Higgs [2/3 of all Higgses; inclusive CMS S / B ∼ 1 / 80] Top tagging – S: large m bb , boost-dependent R bb B: large m bb only for large R bb S/B: ask for large m bb and small R bb boosted Higgs R bb ∼ 2 m H / p T ∼ 0 . 8 ⇒ non-trivial challenge to jet algorthms √ jet definition σ S / fb σ B /fb S / B 30 C/A, R = 1 . 2, MD-F 0.57 0.51 4.4 k ⊥ , R = 1 . 0, y cut 0.19 0.74 1.2 SISCone, R = 0 . 8 0.49 1.33 2.3
Fat jets Higgs tagging Tilman Plehn New strategy for H → bb [Butterworth, Davison, Rubin, Salam; Adam’s and Michael’s talks] Physics case – desperately needed for light Higgs Higgs tagging [2/3 of all Higgses; inclusive CMS S / B ∼ 1 / 80] Top tagging – S: large m bb , boost-dependent R bb B: large m bb only for large R bb S/B: ask for large m bb and small R bb boosted Higgs R bb ∼ 2 m H / p T ∼ 0 . 8 ⇒ non-trivial challenge to jet algorthms Higgs tag in t ¯ tH [TP , Salam, Spannowsky] – uncluster one-by-one: j → j 1 + j 2 1– unbalanced m j 1 > 0 . 8 m j means QCD; discard j 2 2– soft m j 1 < 30 GeV means QCD; keep j 1 – double b tag three leading J = p T , 1 p T , 2 (∆ R 12 ) 4 vs m filt bb no mass constraint — side bin possibly add’t balance criterion – QCD jets everywhere underlying event and pileup deadly filter reconstruction jets [Butterworth–Salam] decay plus one add’l jet at R filt ∼ R jj / 2 reconstruct masses w/ QCD jet – S / B ∼ 1 / 2, so 5 σ for 100 fb − 1 makes sense
Fat jets Higgs tagging Tilman Plehn New strategy for H → bb [Butterworth, Davison, Rubin, Salam; Adam’s and Michael’s talks] Physics case – desperately needed for light Higgs Higgs tagging [2/3 of all Higgses; inclusive CMS S / B ∼ 1 / 80] Top tagging – S: large m bb , boost-dependent R bb B: large m bb only for large R bb S/B: ask for large m bb and small R bb boosted Higgs R bb ∼ 2 m H / p T ∼ 0 . 8 ⇒ non-trivial challenge to jet algorthms Higgs tag in t ¯ tH [TP , Salam, Spannowsky] – uncluster one-by-one: j → j 1 + j 2 1– unbalanced m j 1 > 0 . 8 m j means QCD; discard j 2 2– soft m j 1 < 30 GeV means QCD; keep j 1 – double b tag d σ/ dm b¯ b [fb / 5 GeV] t ¯ tH three leading J = p T , 1 p T , 2 (∆ R 12 ) 4 vs m filt 0 . 6 t ¯ tZ bb t ¯ tjj no mass constraint — side bin 0 . 4 tb ¯ t ¯ b possibly add’t balance criterion 0 . 2 – QCD jets everywhere 0 t ¯ d σ/ dm b¯ b [fb / 5 GeV] tH 0 . 6 underlying event and pileup deadly t ¯ tZ filter reconstruction jets t ¯ tjj [Butterworth–Salam] 0 . 4 t ¯ tb ¯ b decay plus one add’l jet at R filt ∼ R jj / 2 0 . 2 reconstruct masses w/ QCD jet 0 – S / B ∼ 1 / 2, so 5 σ for 100 fb − 1 makes sense 30 60 90 120 150 180 m b¯ [GeV] b
Fat jets Top tagging Tilman Plehn Standard Model top tagger [TP , Salam, Spannowsky, Takeuchi] Physics case Higgs tagging – known for heavy resonances [Johns Hopkins, Stony Brook, Princeton, Washington, Atlas] Top tagging – cool for t ¯ tH : fat Higgs and top jets 0 . 25 t ¯ tH t ¯ tb ¯ – start like Higgs tagger b [but R=1.5] t ¯ tZ 0 . 2 – kinematic selection [filtered] m rec 0 . 15 = 150 ... 200 GeV t m rec W = 60 ... 95 GeV 0 . 1 helicity angle large [learn from single tops] 0 . 05 – no check with side bands 0 – QCD tag typical for signal 60 80 100 120 140 160 180 200 220 m rec m rec W t
Fat jets Top tagging Tilman Plehn Standard Model top tagger [TP , Salam, Spannowsky, Takeuchi] Physics case – known for heavy resonances Higgs tagging [Johns Hopkins, Stony Brook, Princeton, Washington, Atlas] Top tagging – cool for t ¯ tH : fat Higgs and top jets – start like Higgs tagger [but R=1.5] – kinematic selection [filtered] m rec = 150 ... 200 GeV t m rec W = 60 ... 95 GeV helicity angle large [learn from single tops] – no check with side bands – QCD tag typical for signal Results and challenges 5 10 Number of tops – focus on p T < 400 GeV top – O ( 50 %+) efficiency for boosted tops – reconstruct top momentum? inside R=1.5 4 – identify three decay sub-jets? 10 ⇒ give us a little more time tagged 3 10 200 300 400 500 p [GeV] T
Fat jets Top tagging Tilman Plehn Standard Model top tagger [TP , Salam, Spannowsky, Takeuchi] Physics case – known for heavy resonances Higgs tagging [Johns Hopkins, Stony Brook, Princeton, Washington, Atlas] Top tagging – cool for t ¯ tH : fat Higgs and top jets – start like Higgs tagger [but R=1.5] – kinematic selection [filtered] m rec = 150 ... 200 GeV t m rec W = 60 ... 95 GeV helicity angle large [learn from single tops] – no check with side bands – QCD tag typical for signal Results and challenges top inside R=1.5 1 ∈ t – focus on p T < 400 GeV inside fatjet 0.8 – O ( 50 %+) efficiency for boosted tops – reconstruct top momentum? 0.6 tagged – identify three decay sub-jets? 0.4 ⇒ give us a little more time 0.2 tagged with QCD 0 200 300 400 500 p [GeV] T
Fat jets Outlook Tilman Plehn Top and Higgs tagging Physics case Higgs tagging – great success in VH production Top tagging – possible rescue of t ¯ tH production – Higgs tag based on bottom tags and side bin – top tag based on reconstructed masses
Fat jets Tilman Plehn Physics case Higgs tagging Top tagging
Recommend
More recommend