| 1 |
|
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 2 |
|
\section{Introduction}\label{section:Intro} |
| 3 |
|
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 4 |
< |
This note describes an independent search for $H \rightarrow ZZ$ decays in the 4-lepton channel. Currently the analysis is performed in the context of the $1.6~\rm{fb^{-1}}$ LP dataset. We consider two kinematic scenarios, consistent with the direction of the baseline $4\ell$ analysis~\cite{ref:baseline}. The primary differences in this analysis with respect to the current baseline are : |
| 4 |
> |
This note describes a search for $H \rightarrow ZZ^{*}$ in the 4-lepton channel with $4.7~\rm{fb^{-1}}$. The primary differences between this analysis and the current baseline~\cite{baseline} are : |
| 5 |
|
|
| 6 |
|
\begin{itemize} |
| 7 |
– |
\item {\bf MVA Electron Identification :} Electrons are identified using a Boosted Decision Tree discriminant trained on samples of candidates from data. |
| 7 |
|
|
| 8 |
< |
\item {\bf Variable Electron Identification Working Points :} Electron candidates that form the on-shell Z in the 4 lepton system are selected with a looser set of ID requirements relative to those forming the off-shell Z. |
| 8 |
> |
\item {\bf MVA Electron Identification :} Electrons are identified using a Boosted Decision Tree discriminant trained on data. Identification requirements are looser for electron canidates used in the reconstruction of the on-shell $Z$. |
| 9 |
|
|
| 10 |
< |
\item {\bf Anuglar Discrimination :} We include angular information in the signal extraction fit via an MVA discrminant. |
| 12 |
< |
|
| 13 |
< |
\end{itemize} |
| 10 |
> |
\item {\bf 2D Signal Extraction :} We include angular information and the mass of the second (potentially off-shell) $Z$ in the signal extraction fit via an MVA discriminant. |
| 11 |
|
|
| 12 |
+ |
\item {\bf Isolation :} We apply PF isolation criteria on individual leptons. |
| 13 |
|
|
| 14 |
< |
Pay particular attention to jet backgrounds, which are difficult to estimate and can be sneaky. |
| 14 |
> |
\item {\bf Treatment of Fake/Instrumental Backgrounds :} We produce a combined estimate of light and heavy flavor backgrounds using the ``fakeable object'' technique. We assess the performance of this method in several control regions. |
| 15 |
|
|
| 16 |
< |
Our analyses differ on several addtional points ({\it e.g.: } application of lepton isolation, fake and heavy flavor background estimation, {\it etc} ), which we explain in the text that follows. Some aspects of the analysis are not yet complete. The text describes our strategy for these missing pieces. |
| 16 |
> |
\end{itemize} |
| 17 |
|
|
| 18 |
< |
Overall, our changes to event seleation result in {\bf +9/-4 } events in the loose-kinematics signal region. We discusse these events in detail in Section~\ref{section:Results}. The limits we extract are comparable to those of~\cite{ref:baseline} for Higgs masses below $\sim 200~\rm{GeV}$, but improve for higher masses ({\bf Verify}). |
| 18 |
> |
These changes result in the selection of {\bf $+13/-20 = 7$ } events with respect to the baseline analysis. We discuss these events in detail in Section~\ref{sec:Results}. Our observed limits are comparable to those of~\cite{baseline}. |
| 19 |
|
|
| 20 |
|
|
| 21 |
|
|
