RevModPhys.84.1477】Tests of the standard electroweak model at the energy frontier

A signal to noise ratio which is roughly 25 times

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: either W ! e or W ! . The event topology is thus one high-pT charged lepton, ET corresponding to the neutrino, and two or more 6 jets. Events are required to have fired single electron, single muon, electron þ jet, or muon þ jet triggers. Both experiments use events with two or three reconstructed jets, and D0 also makes use of 4-jet events. Table XXV shows the yields Hobbs, Neubauer, and Willenbrock: Tests of the standard electroweak model at . . . 1510 TABLE XXV. The integrated luminosity, predicted signal, backgrounds, and event yields in the W þ jets (b tag) samples used in the CDF and D0 single-top analyses. D0 2:3 fbÀ1 ¼ 3 jets ! 4 jets 139 Æ 18 63 Æ 10 21 Æ 5 1; 829 Æ 161 229 Æ 38 in Z þ jets 222 Æ 35 196 Æ 50 2615 Æ 192 2579 637 Æ 61 85 Æ 17 in Z þ jets 436 Æ 66 73 Æ 17 1; 294 Æ 107 1216 180 Æ 18 26 Æ 7 in Z þ jets 484 Æ 71 30 Æ 6 742 Æ 80 724 R Ldt process ¼ 2 jets tb þ tqb signal tb signal tqb signal W þ jets Z þ jets Diboson " tt Multijet Total predicted Yield 3:2 fbÀ1 ‘ þ jets CDF 77:3 Æ 11:2 113:8 Æ 16:9 1551:0 Æ 472:3 52:1 Æ 8:0 118:4 Æ 12:2 686:1 Æ 99:4 777:9 Æ 103:7 3376:5 Æ 504:9 3315 2:1 fbÀ1 ET þ jets 6 29:6 Æ 3:7 34:5 Æ 6:1 304:4 Æ 115:5 128:6 Æ 53:7 42:1 Æ 6:7 184:5 Æ 30:2 679:4 Æ 27:9 1404 Æ 172 1411 for the D0 and CDF initial selections in which events are required to have at least one jet having a b tag. The final results for both experiments are based on combining results from MV techniques which exploit correlations among variables and event weighting methods based on calculated matrix elements for signal and background events. The methods are applied to events selected by the initial offline requirements. D0 uses three methods: (1) a ME method, (2) a Bayesian neural network (BNN) method, and (3) a boosted decision tree (BDT) method, and the final result comes from the combination of the individual results including correlations. CDF uses five methods: (1) a ME method, (2) a joint likelihood (LF) method, (3) a neural network (NN) method, (4) a BDT method, and (5) projected likelihood functions (LFS), and the final result is the combination of these including correlations among them. Final results are determined using likelihood fits. Likelihoods are created from the multivariate outputs for each lepton flavor, jet multiplicity, and b-tag multiplicity for a given multivariate method. The joint likelihood formed by the product of the individual likelihoods for a given MV method is then maximized as a function of the signal cross sections with systematic correlations between lepton species, jet multiplicity, and b-tag multiplicity taken into account. The BLUE method (Lyons, Gibaut, and Clifford, 1988) is used by D0 to combine the results from each MV method into a single result. The CDF results are combined using a super discriminant (Aaltonen et al., 2008d), in this case an additional neural FIG. 46 (color online). analysis. TABLE XXVI. Single-top cross-section measu...
View Full Document

Ask a homework question - tutors are online