cmsnotes/StopSearch/introbkg.tex


[MAJOR REWORKING OF TEXT: \\
DEFINE PEAK REGION \\
DESCRIBE SCALING PROCEDURE - POSSIBLY ADD EQUATIONS AND SOME \\
INFORMATION FROM OLD SINGLE LEPTON DESCRIPTION CALLED
singlelepbkg\_OLD.tex] 

In order to search for a possible signal from stop decays giving rise to a signature of \ttbar\ with additional \met\
from the LSPs, it is necessary to determine the composition of the SM backgrounds in the signal region. 
This section details the methods pursued to estimate the background in the signal sample and describes the 
procedure to estimate the systematic uncertainties. The general strategy is to use the MC prediction for the 
backgrounds after applying corrections derived from data. 

The most important background to a stop signal arises from SM \ttbar. The \ttbar\ background may be 
separated into contributions containing a single lepton \ttlj\ and two leptons \ttll. As described in this section, 
the \ttll\ background is the dominant process in our signal region (\met\ $>$ 100~\GeV and \mt\ $>$ 150~\GeV, $\ge 1$ b-tags, isolated track veto),
contributing $\sim 80\%$ of the background yield.
%
This background has large true \met\ and consequently larger \mt\ due to the presence of two neutrinos.
Additional contributions to the single lepton sample arise from \wjets\ and single top. The combination of 
all single lepton backgrounds, \ttlj, \wjets\ and single top, comprises $\sim 15\%$ of the signal sample. 
Finally, other background sources such as dibosons, \dy\ + jets, in addition to rarer processes such as \ttbar\ 
produced in association with a vector boson, dibosons and tribosons,
provide a combined contribution to the signal 
sample at the level of $\sim 5\%$.
Finally, the QCD background contribution is small, particularly in the signal sample, with a large \met\ requirement.

The total bkg in the signal region is estimated according to:

$$ N_{bkg}  = N_{1 lep} + N_{2 lep} + N_{rare} $$

$$ N_{1 lep} = N_{1 lep}^{MC} 
\times 
{(1- \epsilon_{fake})^{data} \over (1 - \epsilon_{fake})^{MC}} 
\times 
{N_{peak}^{data} \over N_{peak}^{MC}}
$$

$$ N_{2 lep} = N_{2 lep}^{MC} 
\times 
{(1- \epsilon_{iso\ trk})^{data} \over (1 - \epsilon_{iso\ trk})^{MC}} 
\times 
{N_{peak}^{data} \over N_{peak}^{MC}}
$$

All of these terms will be defined clearly in this section, including
their corrections and sources of systematic errors.

%We then define the following subsamples within this preselection sample:
%\begin{itemize}
%\item $N_{b-tag} = 0$, i.e. b-veto region $\to$ used to validate the lepton + jets bkg estimation method (see Section~\ref{sec:bkg_singlelep}).
%For raw yields prior to any corrections see Tables~\ref{tab:bvetoyieldpeak} and ~\ref{tab:bvetoyieldtail}.
%%
%\item $N_{b-tag} \ge 1 $, i.e. b-tagged region
%For raw yields prior to any corrections see Table~\ref{tab:btagpreselection}.
%%
%\begin{itemize}
%\item with $\met > 100 \GeV$, $ 60 < \mt < 100 \GeV$, and without an additional isolated track veto $\to$ used to normalize top bkg (see Section~\ref{sec:topnorm}).
%For raw yields prior to any corrections see Table~\ref{tab:btagpeakregionnotrkiso}.
%%
%\item with an additional isolated track veto, $\met > 100 \GeV$, $\mt > 150 \GeV$ $\to$ used as signal region 
%For raw yields prior to any corrections see Table~\ref{tab:btagtailaftertrkiso}. As this is our signal region, only MC is shown at this point.
%\end{itemize}
%\end{itemize}

Revision:	1.5
Committed:	Wed Oct 3 05:48:26 2012 UTC (12 years, 7 months ago) by vimartin
Content type:	application/x-tex
Branch:	MAIN
CVS Tags:	HEAD
Changes since 1.4:	+6 -0 lines
Log Message:	updated missing info
#	User	Rev	Content
1	vimartin	1.1
2	vimartin	1.5	[MAJOR REWORKING OF TEXT: \\
3			DEFINE PEAK REGION \\
4			DESCRIBE SCALING PROCEDURE - POSSIBLY ADD EQUATIONS AND SOME \\
5			INFORMATION FROM OLD SINGLE LEPTON DESCRIPTION CALLED
6			singlelepbkg\_OLD.tex]
7
8	vimartin	1.1	In order to search for a possible signal from stop decays giving rise to a signature of \ttbar\ with additional \met\
9	benhoob	1.2	from the LSPs, it is necessary to determine the composition of the SM backgrounds in the signal region.
10	vimartin	1.1	This section details the methods pursued to estimate the background in the signal sample and describes the
11			procedure to estimate the systematic uncertainties. The general strategy is to use the MC prediction for the
12			backgrounds after applying corrections derived from data.
13
14			The most important background to a stop signal arises from SM \ttbar. The \ttbar\ background may be
15			separated into contributions containing a single lepton \ttlj\ and two leptons \ttll. As described in this section,
16	fkw	1.3	the \ttll\ background is the dominant process in our signal region (\met\ $>$ 100~\GeV and \mt\ $>$ 150~\GeV, $\ge 1$ b-tags, isolated track veto),
17			contributing $\sim 80\%$ of the background yield.
18			%
19			This background has large true \met\ and consequently larger \mt\ due to the presence of two neutrinos.
20	vimartin	1.1	Additional contributions to the single lepton sample arise from \wjets\ and single top. The combination of
21			all single lepton backgrounds, \ttlj, \wjets\ and single top, comprises $\sim 15\%$ of the signal sample.
22			Finally, other background sources such as dibosons, \dy\ + jets, in addition to rarer processes such as \ttbar\
23	vimartin	1.4	produced in association with a vector boson, dibosons and tribosons,
24			provide a combined contribution to the signal
25			sample at the level of $\sim 5\%$.
26	benhoob	1.2	Finally, the QCD background contribution is small, particularly in the signal sample, with a large \met\ requirement.
27	vimartin	1.1
28	fkw	1.3	The total bkg in the signal region is estimated according to:
29
30			$$ N_{bkg} = N_{1 lep} + N_{2 lep} + N_{rare} $$
31
32			$$ N_{1 lep} = N_{1 lep}^{MC}
33			\times
34			{(1- \epsilon_{fake})^{data} \over (1 - \epsilon_{fake})^{MC}}
35			\times
36			{N_{peak}^{data} \over N_{peak}^{MC}}
37			$$
38
39			$$ N_{2 lep} = N_{2 lep}^{MC}
40			\times
41			{(1- \epsilon_{iso\ trk})^{data} \over (1 - \epsilon_{iso\ trk})^{MC}}
42			\times
43			{N_{peak}^{data} \over N_{peak}^{MC}}
44			$$
45
46	vimartin	1.4	All of these terms will be defined clearly in this section, including
47			their corrections and sources of systematic errors.
48
49			%We then define the following subsamples within this preselection sample:
50			%\begin{itemize}
51			%\item $N_{b-tag} = 0$, i.e. b-veto region $\to$ used to validate the lepton + jets bkg estimation method (see Section~\ref{sec:bkg_singlelep}).
52			%For raw yields prior to any corrections see Tables~\ref{tab:bvetoyieldpeak} and ~\ref{tab:bvetoyieldtail}.
53			%%
54			%\item $N_{b-tag} \ge 1 $, i.e. b-tagged region
55			%For raw yields prior to any corrections see Table~\ref{tab:btagpreselection}.
56			%%
57			%\begin{itemize}
58			%\item with $\met > 100 \GeV$, $ 60 < \mt < 100 \GeV$, and without an additional isolated track veto $\to$ used to normalize top bkg (see Section~\ref{sec:topnorm}).
59			%For raw yields prior to any corrections see Table~\ref{tab:btagpeakregionnotrkiso}.
60			%%
61			%\item with an additional isolated track veto, $\met > 100 \GeV$, $\mt > 150 \GeV$ $\to$ used as signal region
62			%For raw yields prior to any corrections see Table~\ref{tab:btagtailaftertrkiso}. As this is our signal region, only MC is shown at this point.
63			%\end{itemize}
64			%\end{itemize}
65