Update

trunk/AN-18-026.tex

@@ -24,7 +24,7 @@
 %%%%%%%%%%%%%%%  Title page %%%%%%%%%%%%%%%%%%%%%%%%
 \cmsNoteHeader{AN-18-026} % This is over-written in the CMS environment: useful as preprint no. for export versions
 % >> Title: please make sure that the non-TeX equivalent is in PDFTitle below
-\title{Search for mono-Higgs signatures with $\text{H}\rightarrow\text{b}\bar{\text{b}}$ decays in 2016 data}
+\title{Search for drk Higgs boson production in 2016 data}
 
 % >> Authors
 %Author is always "The CMS Collaboration" for PAS and papers, so author, etc, below will be ignored in those cases
@@ -70,10 +70,10 @@
 % 2. **DO NOT use %**                  to comment out sections of the abstract: the extractor will still grab those lines (and they won't be comments any longer!).
 % 3. For PASs: **DO NOT use tex macros**         in the abstract: CDS MathJax processor used on the abstract doesn't understand them _and_ will only look within $$. The abstracts for papers are hand formatted so macros are okay.
 \abstract{
-We present a search for events with a signature characterized by large
-missing transverse energy and a boosted Higgs boson that has decayed
-into a pair of bottom quarks in pp collisions at a center-of-mass
-energy of 13 TeV in a dataset corresponding to an integrated luminosity of 36\,$\text{fb}^{-1}$. Observations are interpreted in terms of new particles and couplings in various theoretical scenarios that predict such signatures in the context of Beyond Standard Model physics and associated Dark Matter production.  Upper bounds on the production cross sections of such scenarios are placed.
+We present a search for dark Higgs boson production in pp collisions at a center-of-mass energy of 13 TeV in a dataset corresponding to an integrated luminosity of 36\,$\text{fb}^{-1}$. Observations are interpreted in terms of new particles and couplings in various theoretical scenarios that predict such signatures in the context of Beyond Standard Model physics and associated Dark Matter production.  Upper bounds on the production cross sections of such scenarios are placed.
+
+%events with a signature characterized by large missing transverse energy and a boosted Higgs boson that has decayed
+%into a pair of bottom quarks in pp collisions at a center-of-mass energy of 13 TeV in a dataset corresponding to an integrated luminosity of 36\,$\text{fb}^{-1}$. Observations are interpreted in terms of new particles and couplings in various theoretical scenarios that predict such signatures in the context of Beyond Standard Model physics and associated Dark Matter production.  Upper bounds on the production cross sections of such scenarios are placed.
 }
 
 % >> PDF Metadata
@@ -113,7 +113,7 @@
 
 \input{fit.tex} %edit add description on fitting on mass bin
 
-\input{systematics.tex}
+\input{systematics.tex} %edit, remove pt-depedent; trigger eff below 200 remove
 
 \input{results.tex}
 

trunk/fit.tex

@@ -4,7 +4,7 @@ \section{Signal Extraction}
 The results are extracted performing a binned fit (using the Higgs Combine tool (Ref~\cite{COMBINE})) to the missing energy spectrum in bin of soft drop mass [25 GeV $< m_{SD} <$ 75 GeV],[75 GeV $< m_{SD} <$ 100 GeV],[100 GeV $< m_{SD} <$ 150 GeV] and [150 GeV $< m_{SD} <$ 3000 GeV], fitting simultaneously over all the control regions and a signal region.
 For each control regions, the recoil variable $U$ is computed removing either the photon, the muon(s), or the electron(s)from the \MET calculation. Both the normalization and the shape of the $t\bar{t}$, W+jets, and Z+jets background processes are estimated by deriving a scale factor between data and Monte Carlo in bins of recoil. 
 
- The fit is performed tying each \MET bin of the signal region to the same recoil bin of each control region, treating the systematic uncertainties as nuisance parameters. 
+The fit is performed trying each \MET bin of the signal region to the same recoil bin of each control region, treating the systematic uncertainties as nuisance parameters. 
 Both shape and normalization (rate) uncertainties are taken into account. The rate uncertainties arise from luminosity and cross-section uncertainties, as well as uncertainties in the lepton reconstruction in the control regions. The shape uncertainties are derived for the multi control region fit, as well as from intrinsic uncertainties from the reconstructed objects (jet energy scale, resolution, and missing energy). Uncertainties due to the limited MC statistics (bin-by-bin) are also included.
 
 %In addition, the predictions of $Z\rightarrow\nu\nu$ and $W\rightarrow(\ell)\nu$ are constrained in the signal region, allowing the single-lepton $W$ control regions to constrain the $Z\rightarrow\nu\nu$ estimation.
@@ -36,27 +36,44 @@ \section{Signal Extraction}
 The contribution of W+jets production in the $t\bar{t}$ fail control region (enhanced by the anti-double-b requirement) is modeled with data through the transfer factor $R_i^{b\ell_{fail}}$.
 
 
-The expression for the likelihood that is maximized by the fit for a single category is:
+The expression for the likelihood that is maximized by the fit for a single category for both pass and fail the double-b requirement are:
 
 {\scriptsize
 
 \begin{align}
-  \mathcal{L}(\pmb\mu^{W\rightarrow\ell\nu},\pmb\mu^{Z\rightarrow\nu\nu},\pmb\mu^{t\bar{t}},\mu;\pmb\theta) ~= 
+  \mathcal{L}(\pmb\mu^{W\rightarrow\ell\nu},\pmb\mu^{Z\rightarrow\nu\nu},\pmb\mu^{t\bar{t}},\mu;\pmb\theta)_{pass} ~= 
     & \prod_{i\in\text{bins}} \text{Poisson} \left(d_i^{\ell\ell_{pass}}\Big| B_i^{\ell\ell_{pass}}(\pmb\theta) 
                                                    + \frac{\mu^{Z\rightarrow\nu\nu}_i}{R^{\ell\ell_{pass}}_i(\pmb\theta)}\cdot SF^{\mathrm{Z+jets~mis-tag}}_{pass} \right) \nonumber \\
     & \times \prod_{i\in\text{bins}} \text{Poisson} \left(d_i^{b\ell_{pass}}\Big| B_i^{b\ell_{pass}}(\pmb\theta) 
                                                           + \frac{\mu^{t\bar{t}}}{R^{b\ell_{pass}}_i(\pmb\theta)}\cdot SF^{t\bar{t}~\mathrm{mis-tag}}_{pass} \right) \nonumber \\
     & \times \prod_{i\in\text{bins}} \text{Poisson} \left(d_i^{\ell_{pass}}\Big| B_i^{\ell_{pass}}(\pmb\theta) 
                                                           + \frac{\mu^{W\rightarrow(\ell)\nu}_i}{R^{\ell_{pass}}_i(\pmb\theta)}\cdot SF^{\mathrm{W+jets~mis-tag}}_{pass} 
                                                           + \frac{\mu^{t\bar{t}}}{R^{(b)\ell_{pass}}_i(\pmb\theta)}\cdot SF^{t\bar{t}~\mathrm{mis-tag}}_{pass} \right) \nonumber \\
-    & \times \prod_{i\in\text{bins}} \text{Poisson} \left(d_i^{\ell\ell_{fail}}\Big| B_i^{\ell\ell_{fail}}(\pmb\theta)
+
+%    & \times \prod_{i\in\text{bins}} \text{Poisson} \left(d_i^{\ell\ell_{fail}}\Big| B_i^{\ell\ell_{fail}}(\pmb\theta)
+%                                                   + \frac{\mu^{Z\rightarrow\nu\nu}_i}{R^{\ell\ell_{fail}}_i(\pmb\theta)}\cdot SF^{\mathrm{Z+jets~mis-tag}}_{fail} \right) \nonumber \\
+%    & \times \prod_{i\in\text{bins}} \text{Poisson} \left(d_i^{b\ell_{fail}}\Big| B_i^{b\ell_{fail}}(\pmb\theta)
+%                                                          + \frac{\mu^{W\rightarrow(\ell)\nu}_i}{R^{b\ell_{fail}}_i(\pmb\theta)} \cdot SF^{\mathrm{W+jets~mis-tag}}_{fail}
+%                                                          + \frac{\mu^{t\bar{t}}}{R^{b\ell_{fail}}_i(\pmb\theta)} \cdot SF^{t\bar{t}~\mathrm{mis-tag}}_{fail} \right) \nonumber \\
+%    & \times \prod_{i\in\text{bins}} \text{Poisson} \left(d_i^{\ell_{fail}}\Big| B_i^{\ell_{fail}}(\pmb\theta)
+%                                                          + \frac{\mu^{W\rightarrow(\ell)\nu}_i}{R^{(b)\ell_{fail}}_i(\pmb\theta)} \cdot SF^{\mathrm{W+jets~mis-tag}}_{fail} \right)
+    \label{eq:likelihoodPass}
+\end{align}
+
+}
+
+{\scriptsize
+
+\begin{align}
+  \mathcal{L}(\pmb\mu^{W\rightarrow\ell\nu},\pmb\mu^{Z\rightarrow\nu\nu},\pmb\mu^{t\bar{t}},\mu;\pmb\theta)_{fail} ~=
+    & \prod_{i\in\text{bins}} \text{Poisson} \left(d_i^{\ell\ell_{fail}}\Big| B_i^{\ell\ell_{fail}}(\pmb\theta)
                                                    + \frac{\mu^{Z\rightarrow\nu\nu}_i}{R^{\ell\ell_{fail}}_i(\pmb\theta)}\cdot SF^{\mathrm{Z+jets~mis-tag}}_{fail} \right) \nonumber \\
     & \times \prod_{i\in\text{bins}} \text{Poisson} \left(d_i^{b\ell_{fail}}\Big| B_i^{b\ell_{fail}}(\pmb\theta)
-                                                          + \frac{\mu^{W\rightarrow(\ell)\nu}_i}{R^{b\ell_{fail}}_i(\pmb\theta)} \cdot SF^{\mathrm{W+jets~mis-tag}}_{fail}
-                                                          + \frac{\mu^{t\bar{t}}}{R^{b\ell_{fail}}_i(\pmb\theta)} \cdot SF^{t\bar{t}~\mathrm{mis-tag}}_{fail} \right) \nonumber \\
+                                                          + \frac{\mu^{t\bar{t}}}{R^{b\ell_{fail}}_i(\pmb\theta)}\cdot SF^{t\bar{t}~\mathrm{mis-tag}}_{fail} \right) \nonumber \\
     & \times \prod_{i\in\text{bins}} \text{Poisson} \left(d_i^{\ell_{fail}}\Big| B_i^{\ell_{fail}}(\pmb\theta)
-                                                          + \frac{\mu^{W\rightarrow(\ell)\nu}_i}{R^{(b)\ell_{fail}}_i(\pmb\theta)} \cdot SF^{\mathrm{W+jets~mis-tag}}_{fail} \right)
-    \label{eq:likelihood}
+                                                          + \frac{\mu^{W\rightarrow(\ell)\nu}_i}{R^{\ell_{fail}}_i(\pmb\theta)}\cdot SF^{\mathrm{W+jets~mis-tag}}_{fail}
+                                                          + \frac{\mu^{t\bar{t}}}{R^{(b)\ell_{fail}}_i(\pmb\theta)}\cdot SF^{t\bar{t}~\mathrm{mis-tag}}_{fail} \right) \nonumber \\
+    \label{eq:likelihoodFail}
 \end{align}
 
 }

trunk/higgstag.tex

@@ -127,7 +127,7 @@ \subsection{Energy correlation functions}
 The ``pass-over-fail'' in QCD background is also shown in Fig.~\ref{fig:transmap_20percent}. We observe tagging (in)efficiencies that are flat across the entire phase space. 
 
 
-Once a cut on this variable is applied to both MC and data, different residual tagging efficiencies have to be accounted for by means of a scale factor. This scale factor is measured in semileptonic \ttbar events, where the top quark is not fully merged, but one is instead left with a fat jet stemming from a hadronically decaying boosted W boson~\cite{monoH}.
+Once a cut on this variable is applied to both MC and data, different residual tagging efficiencies have to be accounted for by means of a scale factor. This scale factor is measured in semileptonic \ttbar events, where the top quark is not fully merged, but one is instead left with a fat jet stemming from a hadronically decaying boosted W boson~\cite{CMS_AN_2016-219}.
 
 
 %%%
@@ -212,8 +212,8 @@ \subsection{Double-b tagger}
   \label{fig:doublebroc}
 \end{figure}
 
-The chosen working point of double-b $> 0.75$ corresponds to the same background efficiency of the medium working point of the minimum of the CSV scores of the two leading subjets used in the previous iteration of Mono-H  analysis~\cite{monoH}, benefiting from the enhanced signal efficiency.
-In terms of tagging and mistagging efficiency measurements, the analysis is precisely following what has been done in~\cite{CMS-PAS-BTV-15-002} with AK8 jets for deriving scale factors for the double-b tagger to be applied throughout the analysis. The current measurement at 13\TeV with AK8 jets is documented in~\cite{Ref:BTAG2016}. A dedicated tagging and mistagging efficiency mesurements on CA15 jet was derived~\cite{monoH} and will be applied throughout the analysis.
+The chosen working point of double-b $> 0.75$ corresponds to the same background efficiency of the medium working point of the minimum of the CSV scores of the two leading subjets used in the previous iteration of Mono-H  analysis~\cite{CMS_AN_2016-219}, benefiting from the enhanced signal efficiency.
+In terms of tagging and mistagging efficiency measurements, the analysis is precisely following what has been done in~\cite{CMS-PAS-BTV-15-002} with AK8 jets for deriving scale factors for the double-b tagger to be applied throughout the analysis. The current measurement at 13\TeV with AK8 jets is documented in~\cite{Ref:BTAG2016}. A dedicated tagging and mistagging efficiency mesurements on CA15 jet was derived~\cite{CMS_AN_2016-219} and will be applied throughout the analysis.
 
 %\subsubsection{Tagging efficiency measurement}
 %\label{tagging_efficiency}