\documentclass[xcolor=svgnames]{beamer}      
\usepackage[utf8]{inputenc}                  
\usepackage[english]{babel}                  
\usepackage{polski}                          
%\usepackage{amssymb,amsmath}                
%\usepackage[latin1]{inputenc}               
%\usepackage{amsmath}                        
%\newcommand\abs[1]{\left|#1\right|}         
\usepackage{amsmath}                         
\newcommand\abs[1]{\left|#1\right|}          
\usepackage{hepnicenames}                    
\usepackage{hepunits}                        
\usepackage{color}                           







%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%5
\definecolor{mygreen}{cmyk}{0.82,0.11,1,0.25}


\usetheme{Sybila}

\title[Status update o $\color{white}{\Lambda_c \rightarrow p \mu^- \mu^+}$]{Status update o $\color{white}{\Lambda_c \rightarrow p \mu^- \mu^+}$}
\author{Marcin Chrz\k{a}szcz$^{1}$, Tadeusz Lesiak$^{2}$, Mariusz Witek$^{2}$,\\ Borys Nowak$^{2}$}
\institute{$^1$~University of Zurich,\\ $^{2}$ Institute of nuclear Physics}
\date{\today}

\begin{document}
% --------------------------- SLIDE --------------------------------------------
\frame[plain]{\titlepage}
\author{Marcin Chrz\k{a}szcz}
\institute{~(UZH)}

% ------------------------------------------------------------------------------
% --------------------------- SLIDE --------------------------------------------








%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

\section{Motivation}
\begin{frame}\frametitle{Why to search for $\Lambda_c \to \Pproton \mu^{+} \mu^{-}$?}


\begin{itemize}
\item Decay of $\Lambda_c^+ \to \Pproton \mu^+ \mu^-$ is a FCNC.
\item Extremely suppressed in SM due to GIM mechanism. 
\item We will use the experience from $\tau \to \Pproton \mu \mu$.

\end{itemize}

\begin{columns}
\column{2.5in}
\begin{center}
\includegraphics[scale=0.18]{new/FCNC.png}
\end{center}
~$\mathcal{B}(  \Lambda_c^{+} \to p \mu^{-} \mu^{+}  ) < 4.4 \times 10^{-5}$\\
 ~ \@ 90\% CL arXiv:1107.4465 

\column{3.5in}
\includegraphics[scale=0.2]{babar.png}\\
Yield: $11.1 \pm 5.0 \pm 2.5$

\end{columns}

We should easily beat Babar.

	
\end{frame}

\begin{frame}\frametitle{Preliminary selection }

\begin{columns}
\column{2.5in}
~Stripping:
\begin{itemize}
\item $\rm{PID}(\mu)>-5$, $\rm{PID}(\Pproton) >10$
\item $\rm{IPCHi2}>9$, $\rm{PID}(\mu -K)>0$, $\rm{GHOST}<0.3$, $\rm{PID}(\Pproton)>10$, $Pt>300$
\item $\Delta m<150MeV$
\item $c\tau >100\mu m$
\item $\rm{IPChi2} < 225$
\end{itemize}
~Additional:
\begin{itemize}
\item Blind region $\vert m(p\mu\mu) - 2286.46 \vert <40 MeV$.
\item $\phi$, $\omega$ veto.
\end{itemize}
\column{3.5in}

\includegraphics[scale=0.18]{new/Lc_mass.png}\\





\includegraphics[scale=0.2]{new/blind.png}\\

\end{columns}



	
\end{frame}


\section{Strategy}
\begin{frame}

\frametitle{Strategy}
{~}
Follow the strategy of $\tau$ analysis:
\begin{itemize}
\item Take prompt $\Lambda_c$, separate approach to SL.
\item Loose cut preselection.
\item Train MVA on MC prompt signal and recalibrate on data. 
\item Calibrate on date.
\item Normalize to $\Lambda_c^{+} \to \Pproton K^{-} \pi^{+}$, $\Lambda_c^{+} \to  \Pproton \pi^{-} \pi^{+}$ or $\Lambda_c \to \Pproton \phi(\mu \mu)$.
\item Optimise the binning in MVA.
\item CLs method for limit.
\end{itemize}  

\end{frame}

\section{Normalization channel}
\begin{frame}\frametitle{Normalization channel}
\begin{itemize}
\item We have 3 candidates for normalization channel.
\begin{enumerate}
\item  $\Lambda_c \to \Pproton \phi(\mu \mu)$, $BR= (2.4 \pm 0.8) \times 10^{-7} $ 
\item $\Lambda_c^{+} \to \Pproton K^{-} \pi^{+}$, $BR= (5.0 \pm 1.3) \times 10^{-2} $ 
\item $\Lambda_c^{+} \to  \Pproton \pi^{-} \pi^{+}$, $BR= (3.5 \pm 2.0) \times 10^{-3} $
\end{enumerate}
From above list $\Lambda_c \to \Pproton \phi(\mu \mu)$ is a perfect candidate for normalization. 
However Br is a bit low.



\end{itemize}



	
\end{frame}




\begin{frame}\frametitle{Optimising the selection}
\begin{columns}
\column{2.5in}
\begin{itemize}
\item Last time for studies we used BDT that was trained on the fly.
\item Now a student produced a new optimised BDT
\item I include his thesis as attachment to this presentation.  
\end{itemize}


\column{2.5in}
\includegraphics[width=0.95\textwidth]{images/roc.png}\\

\end{columns}


\end{frame}




\begin{frame}\frametitle{Comments about the BDT}

\begin{itemize}
\item From historical reasons we are training this classifier on MC vs MC
\item Problematic part is that we have limited MC background sample.
\item We have how ever the opposite sign (OS) channel: $\PLambdac \to \APproton \APmuon \APmuon$
\item The obvious idea was to use this as an background extrapolation and use it for training and optimisation.
\end{itemize}


\end{frame}

\begin{frame}\frametitle{Differences between SS and opposite sign data}
\begin{columns}
\column{2.5in}
\includegraphics[width=0.95\textwidth]{borys/v_Lambda_cplus_FD_OWNPV_2011_MD.pdf}\\
\includegraphics[width=0.95\textwidth]{borys/v_Lambda_cplus_PT_2011_MD.pdf}\\
\column{2.5in}
\includegraphics[width=0.95\textwidth]{borys/v_mu_two_PT_2011_MD.pdf}\\
\includegraphics[width=0.95\textwidth]{borys/v_proton_PT_2011_MD.pdf}

\end{columns}


\end{frame}

\begin{frame}\frametitle{Normalization channel}
\begin{columns}

\column{2.6in}
\begin{itemize}
\item $\mathcal{O}(500)$ events in pour dataset.
\item Can be used for normalization!
\item With the new BDT we also see small peak of $\Pomega$
\item Will veto that.
\end{itemize}
\column{2.5in}
\includegraphics[scale=0.17]{new/Lc_mass_b.png}\\
\includegraphics[scale=0.17]{new/Lc_mass.png}

\end{columns}



	
\end{frame}


\begin{frame}\frametitle{Possible background}

\begin{center}

    \begin{tabular}{| c | c | c |}
    	\hline
   \textbf{ Resonance} &  $\mathcal{B} (\Lambda_c \to p X)$& $\mathcal{B} (X \to \mu \mu)$\\ \hline 
  
     $\eta$ & UNKNOWN & $(5.8 \pm 0.6) \times 10^{-6}$ \\ \hline
    	 $\rho^0$ & UNKNOWN & $(4.55 \pm 0.28) \times 10^{-5}$ \\ \hline	
	$\omega$ &	UNKNOWN &  $(9.1 \pm 3.0) \times 10^{-5}$ \\ \hline
 	 $f(980)$ & $(2.8 \pm 1.9) \times 10^{-3}$ & UNKNOWN \\ \hline	   
  	 $\phi$ & $(8.2 \pm 2.7) \times 10^{-4} $ & $(2.89 \pm 0.19) \times 10^{-4}$ \\ \hline		     	\hline
  	
   \textbf{ Resonance} & $\mathcal{B} (\Lambda_c \to p X)$ & $\mathcal{B} (X \to \mu \mu \gamma)$\\ \hline 
    $\eta$ & UNKNOWN & $(3.1 \pm 0.4) \times 10^{-4}$ \\ \hline	
    \end{tabular}
\end{center}




	
\end{frame}



\section{Summary}
\begin{frame}\frametitle{Summary}

\begin{itemize}
\item Looks like we will have limits  $\mathcal{O}(10^{-8})$
\item We already see a new $\Lambda_c \to \omega \Pproton$ decay, needs separate analysis
\item Normalization channel is still open, but we are converging towards $\Lambda_c^{+} \to  \Pproton \pi^{-} \pi^{+}$
\item We have one tight cut on the stripping (flight distance), we are considering several solutions.

\end{itemize}


	
\end{frame}


\end{document}