| 1 |
ymaravin |
1.3 |
\section{Cross-checks of the background estimation method}
|
| 2 |
|
|
\subsection{Loosened selection criteria: large statistic check}
|
| 3 |
beaucero |
1.1 |
|
| 4 |
vuko |
1.5 |
As the number of background events after applying the full selection
|
| 5 |
|
|
criteria is small we perform a cross-check of the method with loosened
|
| 6 |
|
|
selection criteria in the $2\mu1e$ signature final state. The loosened
|
| 7 |
|
|
criteria are defined as follows
|
| 8 |
beaucero |
1.1 |
\begin{itemize}
|
| 9 |
vuko |
1.5 |
\item the event must be triggered by the HLTSingleMuonIso trigger,
|
| 10 |
|
|
\item the event must have a $Z\rightarrow \mu^+ \mu^-$ candidate with an invariant mass between 50 GeV and 120 GeV,
|
| 11 |
ymaravin |
1.3 |
\item only one \Z candidate in the event is allowed,
|
| 12 |
ymaravin |
1.4 |
\item a loose electron candidate is required with $p_T > 20$ GeV and $|\eta|<2.5$,
|
| 13 |
|
|
\item an electron and muons candidate must be isolated from each other with $\Delta R>$ 0.1.
|
| 14 |
beaucero |
1.1 |
\end{itemize}
|
| 15 |
|
|
|
| 16 |
ymaravin |
1.3 |
The tight criteria is just an application of the above mentioned requirements with
|
| 17 |
|
|
and additional criterion on the electron candidate to pass tight electron identification
|
| 18 |
|
|
criteria.
|
| 19 |
|
|
|
| 20 |
|
|
Although these criteria enhance the background significantly (see Fig.~\ref{fig:Checkbkg})
|
| 21 |
vuko |
1.5 |
and change the flavor composition of misidentified jets, it can be used to estimate the robustness
|
| 22 |
ymaravin |
1.3 |
of the method to various selection criteria.
|
| 23 |
beaucero |
1.1 |
\begin{figure}[hbt]
|
| 24 |
|
|
\begin{center}
|
| 25 |
|
|
\scalebox{0.8}{\includegraphics{figs/LooseSamples.eps}}
|
| 26 |
ymaravin |
1.3 |
\caption{Invariant mass of the two muon candidates for electron candidate passing loose
|
| 27 |
|
|
criteria (a), and that for electron passing tight criteria (b).}
|
| 28 |
beaucero |
1.1 |
\label{fig:Checkbkg}
|
| 29 |
|
|
\end{center}
|
| 30 |
|
|
\end{figure}
|
| 31 |
|
|
|
| 32 |
ymaravin |
1.3 |
Prior to application of the matrix method described in Section~\ref{sec:D0Matrix}
|
| 33 |
ymaravin |
1.6 |
we subtract \ZZ and \Z$\gamma$ contribution to the di-lepton mass distribution following
|
| 34 |
ymaravin |
1.3 |
exactly the same procedure described
|
| 35 |
|
|
in Section~\ref{sec:SignalExt} for original selection criteria. We also subtract
|
| 36 |
|
|
the background without a genuine \Z boson estimated from fit of the
|
| 37 |
|
|
dimuon invariant mass to a Gaussian convoluted with a Breit-Wigner
|
| 38 |
|
|
function for a \Z mass signal and to a straight line for a background without
|
| 39 |
|
|
a genuine \Z boson.
|
| 40 |
|
|
|
| 41 |
|
|
The comparison between estimated background with the matrix method and the
|
| 42 |
|
|
the sum of the \Z+jets, $W$+jets, $t\bar{t}$+jets and \Z$b\bar{b}$ Monte Carlo truth information
|
| 43 |
|
|
is shown in Fig.~\ref{fig:bkgCheckMatrix} for the loose and the tight sample. Both
|
| 44 |
ymaravin |
1.4 |
predicted and MC truth background distributions agree very well within one $\sigma$ uncertainty,
|
| 45 |
|
|
which indicates the relative robustness of the method to the misidentified jet composition in the
|
| 46 |
|
|
final state.
|
| 47 |
beaucero |
1.1 |
\begin{figure}[hbt]
|
| 48 |
|
|
\begin{center}
|
| 49 |
|
|
\scalebox{0.8}{\includegraphics{figs/MatrixMethodInLooseSamples.eps}}
|
| 50 |
ymaravin |
1.3 |
\caption{Comparison between estimated and MC truth background for a) an electron candidate passing the loose criteria, b) for an electron candidate passing the tight criteria. The errors bars include statistical and
|
| 51 |
|
|
systematic uncertainties estimated from Eq.~\ref{eq:errmatrix}}
|
| 52 |
beaucero |
1.1 |
\label{fig:bkgCheckMatrix}
|
| 53 |
|
|
\end{center}
|
| 54 |
|
|
\end{figure}
|
| 55 |
|
|
|
