• diboson
• fsr
• isr
• technique
• three jets
• WW
• WZ
• ZZ

The study of associated WZ boson production with a lepton and a neutrino signaling the W, and a bb-pair in the final state is important since the event topology of this process is the same as expected for WH associated production of a W and the Standard Model light-Higgs boson (M_{H} < 135 GeV). Thus, the investigation of the process WZ->lnubb whose rate can be accurately predicted, allows to calibrate and optimize many of techniques used in the SM Higgs search and provides a ``standard candle'' for that crucially important search. In addition, WZ associated production generates a significant background for low mass Higgs Boson searches with H decaying into a bb pair.
At the Tevatron, the process WH-> W bb has an expected cross section about five times lower than WZ->Wbb for m_H~ 120 GeV/c2. Therefore, observing that process would be a benchmark for the even more difficult search for the light Higgs in the WH->Wbb process.
Observing associated WZ production at the Tevatron in the channel WZ->lnubb is extremely difficult for two main reasons.
The event rate is extremely low. A WZ production cross section of ~3.22 pb together with a Z->bb branching ratio of ~15% provides about 50 fb in the WZ->lnubb channel. With a trigger and kinematical selection efficiency of the order of a few %, one expects a handful of events per fb^-1 of integrated luminosity.
This statement remains valid even if the few ZZ events with leptonic decay of one Z are included in the acceptance.
A standard kinematical cut requests exactly two high energy jets (i.e. E_{T}> 20 GeV) in the candidate sample. Simulations show that if a third energetic jet is allowed the signal acceptance is increased by about 1/3. Therefore, it would be very important to be able to extract a Z->bb signal also in events with more than two high energy jets.
A second difficulty is that the signal to background ratio is very poor, due primarily to the contribution of associated production of $W$ and incoherent jets. Optimal dijet mass resolution is of utmost importance for discriminating this background, since a fit to the invariant mass distribution of the two jets, associated to the hadronic decay of $Z$, is used to disentangle the diboson signal from the backgrounds in the candidate data sample.
In this thesis, we present a search for WZ/ZZ in events with a lepton(s), missing transverse energy and jets. Besides looking at the sample where two exclusive jets are found, we investigate the sample with 3 jets where about the 33% of the signal events lie.
In WZ events, additional jets may be initiated by gluon(s) radiated by the interacting partons (initial state radiation, ISR) or by the Z-decay products (final state radiation, FSR). FSR jets should legitimately be included in the reconstructed Z-mass.
However, the presence of either ISR or FSR jets in a 3-jets events confuses the choice of the jet system to be attributed to Z decay. In this sample the invariant mass of the two E_T-leading jets would normally be chosen to reconstruct the Z boson. To improve both the mass resolution and the sensitivity of the search we describe an alternative procedure to reconstruct the Z-invariant mass. Improving the resolution in such a sample means choosing the correct jet combination for building the Z mass.
My thesis work has been to investigate at generator level a sample of simulated CDF WZ events for finding a means to determine the origin of the extra jet and the right jet combination to be chosen for the best reconstruction of the Z mass. This is attempted for the first time in CDF.
Four different Neural Networks (NNs) have been trained: NN$_{12}$, NN$_{13}$, NN$_{23}$ and NN$_{123}$. These NNs should make us to be able to decide event by event which among of the 4 four possible different combinations can be used for building the $Z$-mass in the three jets sample. If one jet is due to ISR, we expect one of the MJ1J2, MJ1J3, MJ2J3 combinations to be correct, while if one jet is due to FSR the choice should be MJ1J2J3. NNs combine kinematical information and some tools developed by CDF Collaboration for distinguishing gluon-like and b-like jets from light-flavored jets.
Based on the response of the four NNs, we determine the most likely jet combination for building the Z mass in each event. The method allow to use a different combination from J1J2 in about 50% of cases.
To qualify the potential of the method we have studied an experimental data sample accepting events with a leptonically decaying W and 3 large transverse momentum jets, as in the studies of the simulated WZ sample. The selection cuts accept jets of all flavors (pretag sample), and all diboson events including WW besides WZ, ZZ may pass the cuts. We estimate the probability at three standard deviations level to extract an inclusive diboson signal in the 3-jets sample alone (P$3\sigma$). After our procedure for building the Z mass is applied, P$3\sigma$ is about 4 times greater than when building the Z mass ``by default'' with the two E_{T} leading jets.
The next step would be to discriminate against the WW contribution. A straight ``Higgs like'' approach would be to require b-jets in the events. When one or more jets are required to be b-like (the \emph{tag} sample) our technique, if applied stand-alone, provides only a modest improvement in sensitivity over the option of building the Z-mass from J1J2. Studies for improving the method further are on-going.
However, already now our technique allows including the three jets sample in the WZ/ZZ analyses in order to increase acceptance and sensitivity in the search for the hadronically decaying Z-boson.