TOP MASS AND JET ENERGY SCALE MEASUREMENTS IN THE ALL HADRONIC CHANNEL WITH THE TEMPLATE METHOD (TMT2D) ---- with Neural Net Selection ---- ***** 2.1 fb-1 *****

Authors Luca Brigliadori (Universita' di Trento) Andrea Castro (Universita' di Bologna) Fabrizio Margaroli (Purdue University)

Introduction

We look at t tbar events where both top quarks decay into Wb, and the W's decay to two quarks which give rise to two jets. This channel has the advantage of the largest branching ratio (~44%) and of fully reconstructed kinematics. The main improvements with respect to the previous template analysis based on 1fb-1 of data are the reduction of the systematic effects, most notably those due to the jet energy scale, by taking advantage of the reconstruction of the mass of the 2 hadronically decaying W's. A less significant, but formally important, improvement is the background description, which is now event-based and no more tag-based.

Dataset
We are using the data selected by a multi-jet trigger (2.1 fb-1). For the signal we are using mainly Monte Carlo tt events generated with Pythia and with the top mass between 160 and 190 GeV/c2. Other Monte Carlo samples generated with Pythia and Herwig under different conditions are used to study the systematic uncertainties on the mass measurement.

Event selection (learn more from XSEC web page)

For a minimal clean-up of the samples, we apply the follow pre-requisites: A) number of tight lepton = 0 (to guarantee the orthogonality with lepton+jets analysis); B) |Zjvert| < 60 cm; C) |Zjvert - Zpvert| < 5 cm; D) N(vert_class12) >= 1 E) MET_significance < 3 (to guarantee the orthogonality with tau+jets analysis). F) 6 <= Njet <= 8; (jets with Et>=15, |eta|<=2.0) G) DeltaRmin >=0.5 between jets; A strong reduction of the background is provided by a Neural Net Kinematical selection based on 11 Input variables+ 2 hidden nodes (20 and 10 nodes) The input variables are: 1) SumEt; 2) SumEt3; 3) Aplanarity 4) Centrality; 5) M2jMin; 6) M2jMax; 7) M3jMin; 8) M3jMax; 9) Et1Star(=Et*sin^2\theta^*); 10) Et2Star; 11) <EtStar>3N(geometric average over the 3rd-4th...Nth jets); All of these quantities are based on corrected jets, so they are re-calculated every time we vary the JES. We cut as in the previous analysis requiring the Neural network output to be greater than 0.91. We finally subdivide our data sample in two non-overlapping samples according to their b-tag content: events with exactly one SECVTX tagged jet or exactly two SECVTX tagged jets. NNoutput distribution for 1 tag events, zooming in the signal region (NNout>0.91) (eps) NNoutput distribution for 2 tag events, zooming in the signal region (NNout>0.91) (eps)

Kinematic fitter

For events with 6-to-8 jets we consider the 6 leading jets (highest in Et) and define a chi2 containing 2 dijet masses (set to be equal to the W mass), 2 triplet masses (set to be equal one to the other) and 6 terms representing the jet energy resolution. For each event with at least one b-tagged jet we consider all possible combinations where the tagged jets are assigned to b partons and keep the mass corresponding to the combination with the smallest chi2.

Data-driven background with a 3-dim Matrix (learn more from XSEC web page)

This matrix is made starting from a sample with only 4 jets and using: 1) 6 bins in Number of primary vertices 2) 13 bins in Number of good tracks inside jets (with silicon hits) 3) 5 bins in Et This matrix is used to obtain a prediction of the invariant mass distribution for background events, by weighting each event with the total tagging probability. A specific procedure is applied to account, on average, for correlations among tags. The agreement between expected and observed tagged events in a control region is quite good. Top reconstructed mass for single-tagged events (NN control region) (eps) Top reconstructed mass for double-tagged events (NN control region) (eps) W reconstructed mass for single-tagged events (NN control region) (eps) W reconstructed mass for double-tagged events (NN control region) (eps)

Template parametrization The signal and background templates are parametrized with combinations of one/two gaussian and one/two gamma functions. For the signal templates the parameters depend linearly on the top mass and on the JES. For the background, the expected contribution from tt events is accounted for. Mt Signal templates(1 tag) vs Mtop (eps) Mt Signal templates(2 tags) vs Mtop (eps) Mt Signal templates(1 tag) vs JES (eps) Mt Signal templates(2 tags) vs JES (eps) Mw Signal templates (1 tag) vs Mtop (eps) Mw Signal templates (2 tags) vs Mtop (eps) Mw Signal templates (1 tag) vs JES (eps) Mw Signal templates (2 tags) vs JES (eps) Mt Background template (1 tag) (eps) Mt Background template (2 tags) (eps) Mw Background template (1 tag) (eps) Mw Background template (2 tags) (eps)

Likelihood fit

We define a likelihood function which depends on the number of signal and background events and on the corresponding probability density functions (templates) We then find the input top mass and JES value which maximize the likelihood. Before applying this to the data we test the performance on a set of pseudo-experiments where we sum background and signal events (for different input top masses) extracted from the templates in the expected proportion. This procedure is used for the calibration of the response functions and to estimate possible biases in the measurement. We study the linearity of the values returned by the fit with respect to the input values, the residuals and pulls of these values, and we account for these small biases.

Mass Linearity plot (eps) JES Linearity plot (eps) Mtop Pull width (eps) JES Pull width (eps) Mtop Pull mean (eps) JES Pull mean (eps)

The 2.1 fb-1 top mass measurement
The 2.1 fb-1 top mass measured in the all hadronic channel is: mt = 176.9 +- 3.8 (stat+JES) +- 1.7 (syst) GeV/c2 JES vs Mtop contour (eps) Mtop Expected statistical uncertainty (eps) Mtop: data vs fitted distributions (1 tag) (eps) Mtop: data vs fitted distributions (2 tags) (eps) MW: data vs fitted distributions (1 tag) (eps) MW: data vs fitted distributions (2 tags) (eps)

Systematic uncertainties
Table 1 Source Mtop Systematics (GeV/c2) Residual Bias 0.32 2D calibration 0.03 Generator 0.48 ISR/FSR 0.31 Bkg template 1.03 Sig template 0.26 B-JES 0.55 Btag SF 0.35 Res JES 0.80 PDF 0.41 Mt-MW correl. 0.21 Pile-up 0.28 Total 1.70 Note: some of the uncertainties are very small so we quote the uncertainty on their values to be conservative.