Search for neutral Higgs bosons decaying to tau pairs in association with b-quarks at the D0 Detector Page: 2 of 4
This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to Digital Library by the UNT Libraries Government Documents Department.
The following text was automatically extracted from the image on this page using optical character recognition software:
a neural network (NNT) to distinguish hadronic tau decays from jet fakes. DO has three
different types of hadronic taus, differing in their number of charged tracks and/or the presence
of electromagnetic energy in the calorimeter. We make the following kinematic and neural
network requirements on the hadronic tau candidate:
" Type 1: ET> 10 GeV/c, ptk > 7 GeV/c, NNT> 0.9
" Type 2: ET> 10 GeV/c, ptk > 5 GeV/c, NNT> 0.9
" Type 3: ET> 15 GeV/c, 1 track with pT> 5 GeV/c, Z pk > 10 GeV/c, NNT> 0.95.
Additionally we require at least one jet with pT> 15 GeV/c, irf < 2.5. At least one of the
jets must be tagged as coming from b-quark fragmentation by the DO neural network b-tagger.
Figure 1 shows the data/background comparison in the visible mass (defined as the invariant
mass of the muon, hadronic tau, and missing transverse momentum) before and after b-tagging.
Backgrounds in this search include W/Z+jets production, multijet production, Di-boson,
thand single top production. The W/Z+jets contribution is estimated using ALPGEN Monte
Carlo (MC) interfaced with PYTHIA for hadronization and showering. Signal, Di-boson
production and tE production are estimated using PYTHIA, and single top production via
COMPHEP[7, 8, 9]. Multijet background is estimated primarily from data. We use two
independent methods to estimate multijet production; the first method relies on measuring the
probability for a jet to be b-tagged in a multijet-enriched sample, while the second method uses
the probability for a jet to pass the NNT cut and for a muon to pass the isolation requirement
in a multijet-enriched sample. We take the average of the two methods as the final multijet
contribution and include the difference between the two methods as a systematic error.
D0 Runll Preliminary, 1.2 fb D0 Runll Preliminary, 1.2 fb
tl inc. Il Tau Types tt - incl
70 - WW/WZ - QCD
1 Single Top 10 W Other BKGD
60 - Z+j- +j - Data
50 - ''Zb+ pbb+j- _ Signal M=120 GeV
W+H F+j-ilv H F+j
3 - Z+j-mz+j
30 Z+H F+j-zb+H F+
20 - " Data
10 All Tau Types
00 50 100 150 200 250 0 50 100 150 200 250 300
M MET (0eV) M,,MET (GeV)
Figure 1. Data/background comparison in the visible mass variable (M(p, T, MET) after muon,
hadronic tau, and jet selection. Left: before b-tagging. Right: after b-tagging. In the legend,
"QCD" refers to the multijet background.
3. Multivariate Methods
After b-tagging our data sample is dominated by tt and multijet events. We employ two
multivariate techniques to reject each of the two leading backgrounds. To reject tt background
we apply a Kinematic Neural Network (KNN) originally developed in . It uses the number of
jets in the events, the sum of the transverse momenta of the jets (HT), the energy from the four-
momentum sum of the muon, tau and jets, and the A4 between the muon and tau candidate as
input variables. A KNN cut of 0.3 typically offers ~ 75% rejection in tt with only a ~ 4% signal
loss. To reject QCD we apply a simple unbinned log-likelihood ratio, trained separately for each
Here’s what’s next.
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
Herner, Kenneth. Search for neutral Higgs bosons decaying to tau pairs in association with b-quarks at the D0 Detector, article, June 1, 2009; Batavia, Illinois. (https://digital.library.unt.edu/ark:/67531/metadc929828/m1/2/: accessed April 24, 2019), University of North Texas Libraries, Digital Library, https://digital.library.unt.edu; crediting UNT Libraries Government Documents Department.