EGamma_ElectronTrackingValidation/docs/presentations/2017_08_08/main.tex

% rubber: module pdftex

\documentclass[english,aspectratio=43,9pt]{beamer}
\usepackage{graphicx}
\usepackage{amssymb}
\usepackage{booktabs}
\usepackage{siunitx}
\usepackage{subcaption}
\usepackage{marvosym}
\usepackage{verbatim}

\newcommand{\pb}{\si{\pico\barn}}%
\newcommand{\fb}{\si{\femto\barn}}%
\newcommand{\invfb}{\si{\per\femto\barn}}
\newcommand{\GeV}{\si{\giga\electronvolt}}

\hypersetup{colorlinks=true,urlcolor=blue}

\usetheme[]{bjeldbak}

\newcommand\SmallFont{\fontsize{6}{7.2}\selectfont}

\begin{document}

\title[EGM Recom. August 2017]{EGM Recommendations, August 2017}
\author[C. Fangmeier]{\textbf{Caleb Fangmeier} \\ Georgios Krintiras \\ Till Arndt}
% \institute[UNL]{University of Nebraska \-- Lincoln}
\date{August 8, 2017}

\titlegraphic{%
\begin{figure}
  \vspace{-0.5cm}
  \includegraphics[width=0.75in]{CMSlogo.png} \\
  \vspace{0.3cm}
  \includegraphics[width=0.75in]{nebraska-n.png}\hspace{0.25in}
  \includegraphics[width=0.75in]{UCLlogo.png}\hspace{0.25in}
  \includegraphics[width=0.75in]{DESYlogo.png}
\end{figure}
}


\begin{frame}[plain]
  \titlepage%
\end{frame}


\begin{frame}{Introduction}
  \begin{itemize}
    \item EGM issues recommendations for
      \begin{itemize}
        \item Energy Calculation (Regression)
        \item Energy scale factors and smearing for MC
        \item Identification, both cut-based and MVA
        \item Efficiency scale factors
      \end{itemize}
    \item For 2016 Data, ``Ultimate'' recommendations are available.
    \item For 2017 Data, there are initial recommendations \alert{but many have not yet converged.}
    \item As this is just a 10 min talk, I'll mostly just be pointing to various talks/TWikis with more information.
  \end{itemize}
\end{frame}

\begin{frame}{Energy Regression, \& MC Energy Scale Correction \& Smearing}
  \begin{itemize}
    \item Latest recommendation for 2016 data is still the Moriond 2017 recommendation. (See TWiki)
    \item Still unavailable for 2017 data, stay tuned for announcements.
    \item In mean time, recommendation for 2017 data is to use electrons and photons from standard collections.
    \item \url{twiki.cern.ch/twiki/bin/viewauth/CMS/EGMRegression}
    \item \url{twiki.cern.ch/twiki/bin/viewauth/CMS/EGMSmearer}
  \end{itemize}
\end{frame}

\begin{frame}{Electron Identification \-- Cut Based}
  \begin{columns} \begin{column}{0.5\textwidth}
    \begin{itemize}
        \item Four available working points
          \begin{itemize}
              \item Veto (eff. $\sim 95\%$)
              \item Loose (eff. $\sim 90\%$)
              \item Medium (eff. $\sim 80\%$)
              \item Tight (eff. $\sim 70\%$)
          \end{itemize}
        \item The 80X-tuned ID is recommended for 2016 data.
        \item For now, the same holds for 2017 data. New IDs will be tuned when more data is available. (expected by late fall)
        \item See Ilya's \href{https://indico.cern.ch/event/647765/contributions/2632811/attachments/1480145/2294932/talk_egmIDs_on_90X.pdf}{talk} for more information on 2016 IDs with 2017 MC.
        \item {\footnotesize \url{twiki.cern.ch/twiki/bin/view/CMS/CutBasedElectronIdentificationRun2}}
    \end{itemize}
  \end{column} \begin{column}{0.5\textwidth}
    \begin{figure}
      \vspace{-.5in}
      \includegraphics[width=\textwidth]{figures/electron_cuts.png}
    \end{figure}
  \end{column} \end{columns}
\end{frame}

\begin{frame}{Electron Identification \-- MVA}
  \begin{itemize}
      \item For most analyses, the ``General Purpose MVA'' is recommended.
      \item There is also a more specialized ``HZZ MVA'' developed for the ICHEP 2016 HZZ result specialized for running at very high efficiencies. ($\sim 98\%$)
      \item Two specified working points
        \begin{itemize}
          \item WP80 (eff. $\sim 80\%$)
          \item WP90 (eff. $\sim 90\%$)
        \end{itemize}
    \item Tested on 2017 MC samples and shows reasonable performance,
    \item However, stated WP efficiencies tend to be shifted along the ROC curve.
    \item {\footnotesize \url{twiki.cern.ch/twiki/bin/view/CMS/MultivariateElectronIdentificationRun2}}
  \end{itemize}
  \begin{figure}
    \includegraphics[width=0.75\textwidth]{figures/MVA_performance.png}
  \end{figure}
\end{frame}

\begin{frame}{Electron Identification \-- HEEP (high energy)}
  \begin{itemize}
      \item Specialized ID for very high energy ($\sim 1$TeV) electrons.
    \item As of August 1, HEEP V70 is the recommended version
    \item Implemented in the VID (versioned identification) Framework. See TWiki for instructions on usage/integration.
    \item \url{twiki.cern.ch/twiki/bin/view/CMS/HEEPElectronIdentificationRun2}
  \end{itemize}
\end{frame}

\begin{frame}{Photon Identification}
  \begin{itemize}
      \item Recommendations are similar to electrons.
      \item See relevant TWiki pages for details.
        \begin{itemize}
            \item \url{twiki.cern.ch/twiki/bin/view/CMS/CutBasedPhotonIdentificationRun2}
            \item \url{twiki.cern.ch/twiki/bin/view/CMS/MultivariatePhotonIdentificationRun2}
        \end{itemize}
  \end{itemize}
  \vspace{0.8in}
  \centering
  \alert{Starting in \texttt{9\_2\_2}, all 2016 IDs (for electrons \emph{and} photons) are in CMSSW (no private branch merges!)}
\end{frame}

\begin{frame}{Efficiency and Scale Factors}
  \begin{itemize}
    \item Available for 2016 data for all mentioned IDs.
    \item Moriond 2017 recommendations are most current for 2016 data.
    \item For 2017 data, scale factors will likely be different. EGM will publish updated numbers shortly after 2017 IDs are finalized.
    \item Instructions: \url{twiki.cern.ch/twiki/bin/viewauth/CMS/EgammaIDRecipesRun2} \\ Under \textbf{Electron/Photon efficiencies and scale factors}
    \item More Details: \url{twiki.cern.ch/twiki/bin/view/CMS/ElectronScaleFactorsRun2}
  \end{itemize}
\end{frame}

\begin{frame}{Summary}
  \begin{itemize}
    \item For now, keep using 2016 recipes and constants (fully in CMSSW since \texttt{9\_2\_2}).
    \item \emph{Be mindful that oddities are possible since these have not been exercised on real data so far}.
    \item New recommendations are expected late fall once sufficient data for Z calibrations are collected
    \item Any questions can be directed new contact Caleb Fangmeier \url{caleb@fangmeier.tech}.
  \end{itemize}
\end{frame}


\end{document}