Searches for ttH (H->bb)

This page contains approved plots and results in the order as they appear in the CSC note. Only the CSC note contains all the relevant information and should thus be consulted if one of the plots is used.

Figure 1: One of the Feynman diagrams for ttH production in the semi-leptonic final state.

Figure 2: Example of Feynman diagrams for the ttbb QCD production.

Figure 3: Example of Feynman diagrams for the ttbb EW production.

Figure 4: Left: Multiplicity of jets, inside p_T and η acceptance. Right: Distribution of b-tagging weight for b-, c- and light jets in ttH events, using the IP3D+SV1 tagger.

Figure 5: Rejection of light and c-jets versus b-tagging efficiency. Red open markers indicate the working point (b-tag weight >= 5.5) used for the cut-based and paring likelihood analyses, open markers with a central dot (right plots) represent the performance when the 30% performance degradation is applied. The lower plots show results for purified jets, where no heavier quark existed within a wide cone of ΔR < 0.8.

Figure 6: Effect of adding a reconstructed low p_T muon to at least one of the b jets (left), and the effect of jet calibration (right) shown for the Higgs boson mass where the correct jet combination is chosen. Red solid (black open) markers and solid (dotted) line show the mass distribution and fitted values for jets after (before) correction. The effect of the accidental wrong muon matches can be seen in the shape distortion. All distributions are normalized to unity.

Figure 7: Cross-sections of hadronically decaying W combinations per event giving candidates within 25 GeV of the true W mass in the signal sample. (left). Right: Invariant mass spectrum for hadronically decaying W candidates normalized to unity. The dotted line shows combinations where the jets from the W are correctly matched.

Figure 8: Distributions of ΔR between the true and the reconstructed W boson (left) and of the reconstructed leptonically decaying W mass (right) for events where a solution for pz is found (solid black line) and events where an approximation is used (dotted red for Δ = 0, dashed blue for the collinear approximation). All distributions are normalized to unity.

Figure 9: Reconstructed invariant mass spectrum for selected leptonic (left) and hadronic (right) top quark candidates in the signal sample. The dotted red line indicates the candidates formed by assigning the correct b-jet to the top quark being considered. All distributions are given in cross-section.

Figure 10: Left: Reconstructed invariant mass spectrum for Higgs boson candidates in the signal sample. The dotted red line indicates the candidates formed by assigning the correct b-jets. Right: Reconstructed invariant mass spectrum for signal and backgrounds after the cut-based selection. All distributions are given in cross-section.

Figure 11: Pairing likelihood templates for top quark topological distributions, derived from the ttH signal sample. Solid lines represent the correct combination while the dotted lines show the combinatorial background in the signal itself. See text for a description of the variables.

Figure 12: Left hand side: combinatorial likelihood output for ttH events. Black solid (red dotted) histogram indicates the correct (wrong) combinations. Right hand side: invariant mass for the Higgs boson candidates reconstructed using the maximum likelihood configuration, after applying a cut on the likelihood. Dotted histogram indicates the correct combinations. The differential cross-section is shown in fb.

Figure 13: On the left (right) hand side is shown the leptonic (hadronic) top quark candidates reconstructed invariant mass using the maximum likelihood configuration, after applying a cut on the likelihood output. The dotted histogram indicates the correct b quark jet for the top quark being considered. The differential cross-section is shown in fb.

Figure 14: Reconstructed invariant mass spectrum for Higgs boson candidates for signal and backgrounds after pairing likelihood selection. The differential cross-section is shown in fb.

Figure 15: The b-tagging likelihood ratio extracted from signal simulation as the ratio of b-weight distributions for b-jets and light jets. The degree of smoothing reflects the statistical precision at each point.

Figure 16: Left: The jet charge of W boson candidates, based on the sum of the jet-charges of the two jets signed by the high-p_T lepton. Center: Jet charges of individual b-jets from top quarks; signed so that those where quark charge agreeing with the lepton charge from W decay are positive. Right: The magnitude of the sum of the jet charges of the jets assigned to the Higgs boson. Wrong b-jet combinations have a somewhat flatter distribution. Correct combinations are solid (black) and wrong combinations are dashed (red).

Figure 17: Left: The χ² of the fit. Center: jet-top|_min . The cosine of the minimum angle between top quarks and their daughter jets. Correct combinations have more collimated tops. Right: The t_b^leptcosΘ^*. The decay angle in the rest frame of the leptonically decaying top quark of the b-jet relative to the top quark direction in the lab frame. Correct combinations are solid (black) and wrong combinations are dashed (red).

Figure 18: The distribution of distances, ΔR, between the lepton and the b quark from the leptonic top quark. Center: The maximum |η| of any jet in the combination being tested. Correct combinations are more central. Right: The reconstructed mass of the hadronically decaying top quark, before any fit is performed. Correct combinations are solid (black) and wrong combinations are dashed (red). Caption Text

Figure 19: The performance of a range of possible variables if they are used individually to find the correct quark-jet pairing in a ttH event. For a given efficiency for selecting the correct pairing (x axis), what fraction of the incorrect pairings will also be chosen (y axis). The line labelled null shows the effect of selecting combinations at random.

Figure 20: Left: Distributions for signal and backgrounds of the likelihood used to find the combinatorics solution. Center: The sum of the b-tag likelihoods of the four jets used as b's in the chosen combination. There was cut at 8 in the combinatorics preselection. Right: The sum of the b-tags of the jets associated to the Higgs boson. This variable removes ttjj more than ttbb. Histograms are filled for every event where there is a successful fit. In all histograms, the sum of the individual histograms is shown. They are stacked to indicate relative contributions.

Figure 21: Left: The maximum cosine of the angle between either top quark and the Higgs boson when boosted into the ttH center of mass frame. Center: The momentum of the Higgs boson candidate in the center of mass frame. Note that the true Higgs bosons have a lower momentum than the background. Right: The cos Θ^* of the higher p_T of the jets from the Higgs boson. This tends to be more central for signal events. In all histograms, the sum of the individual histograms is shown. They are stacked to indicate relative contributions.

Figure 22: Left: The total signal separation likelihood. The top figure shows all events while the bottom shows only those within a Higgs boson candidate mass window of 90 to 150 GeV. Right: The integrated version of the lower left plot, so the total event rates passing any cut can be seen. The bottom half of this plot is the signal to background ratio implied.

Figure 23: The mass distribution after cutting at -4.2 in Ls/b . Left: All samples, showing the contributions stacked. The signal distribution is also shown separately at the bottom. Right: Signal only, the dashed (red) line equals events where the correct jets from the Higgs boson are selected.

Figure 24: Comparison of the purity of reconstructed bb invariant mass before the final mass window cut versus selection efficiency for the cut-based, the pairing likelihood and the constrained fit analysis. The solid markers show the selected working points.

Figure 25: Ratio of the invariant mass spectrum for Higgs boson candidates after combinatorial likelihood analysis and using a loose and tight cut on the b-tag weight. Left hand side: tt+jets, right hand side: ttbb. The signal region shows very consistent behaviour.

Figure 26: High p_T lepton multiplicity (e^+/-, μ^+/- ) for pile-up and non pile-up samples (before preselection).

Figure 27: Jet multiplicity (before preselection) for pile-up and non pile-up samples. The cuts p_T < 20 GeV and |η| < 5 are applied to the individual jets.

Figure 28: b jet multiplicity (after preselection) for pile-up and non pile-up samples. The cuts p_T < 20 GeV, |η| < 2.5 and b jet weight > 5.5 are applied to the individual jets.

Figure 29: Reconstructed Higgs boson mass peak (mbb ) for pile-up and non pile-up samples.

Figure 30: Comparison of the total significance as function of systematic uncertainties (ΔB), for the cut-based, the pairing likelihood and the constrained fit analysis. Markers indicate the significance corresponding to the background uncertainty estimated in Table 8.