main.tex

\documentclass[11pt]{article}

% --- Packages ---
\usepackage[usenames, dvipsnames]{color} % Cool colors
\usepackage{enumerate, amsmath, amsthm, amssymb, mathrsfs, algorithm, algpseudocode, fontawesome, pifont, subfig, fullpage, csquotes, dashrule, tikz, bbm, booktabs, bm, hyperref, wasysym}
\usepackage[framemethod=TikZ]{mdframed}
\usepackage[numbers]{natbib}
\usepackage[normalem]{ulem}

% --- Misc. ---
\hbadness=10000 % No "underfull hbox" messages.
\setlength{\parindent}{0pt} % Removes all indentation.

% -- Commands --
\input{commands}

% Nice coloring of references.
\usepackage{hyperref}
\hypersetup{               
    colorlinks=true,              
    breaklinks=true,               
    urlcolor=dblue,
    linkcolor=dgreen,
    linkbordercolor=dgreen,
    citecolor=dblue,
    citebordercolor=dblue
}



\title{ICML 2019 Notes \\ \Large{Long Beach, CA, USA}}
\author{David Abel\footnote{\durl{http://david-abel.github.io}} \\ \durl{david_abel@brown.edu}}
\date{June 2019}

\begin{document}
\maketitle
\tableofcontents
\newpage


This document contains notes I took during the events I managed to make it to at ICML, in Long Beach, CA, USA. Please feel free to distribute it and shoot me an email at \durl{david_abel@brown.edu} if you find any typos or other items that need correcting. 


\section{Conference Highlights}


I spent most of my time at the RL sessions this round (and sadly missed all of the keynotes), so most of my reflections (and notes) are concentrated on RL:
\begin{enumerate}
    \item Lots of great work on off-policy evaluation and off-policy learning (see, for instance, work by \citet{hanna2018importance,le2019batch,fujimoto2018off,gottesman2019combining}, and talks in Section~\ref{sec:off_pol}). These problem settings are really important, as I (and many others) anticipate RL applications will come along with loads of data from sub-optimal policies.
    
    \item Exploration was a hot topic again, and rightfully so (see work by~\citet{mavrin2019distributional,fatemi2019dead,hazan2018provably,shani2019exploration}). Along with off-policy evaluation (and a few others), it's one of the foundational problems in RL that we're in a good position to make serious progress on at the moment.
    
    \item Some really nice work continuing to clarify distributional RL~\cite{bellemare2017distributional} (see work by~\cite{rowland2019statistics,mavrin2019distributional,qu2018nonlinear}).
    
    \item The AI for climate change workshop on Friday was fantastic and extremely well attended (standing room only for the talks I was there for). I've said this after previous conferences, but: as we all know, there are profoundly important problems, and the tools of ML can be extremely effective in their current form.
    
    \item I really think we need to standardize evaluation in RL. Not that we only need a single method for doing so, or a single domain, but at the moment there is far too much variance in evaluation protocols.
    
    \item Loved the panel at the RL for real life workshop (see Section~\ref{sec:panel})
\end{enumerate}



% ------------
% -- Sunday --
% ------------
\newpage
\section{Monday June 10th: Tutorials}
\input{days/monday_6_10_19.tex}



% ------------
% -- Monday --
% ------------
\newpage
\section{Tuesday June 11th: Main Conference}
\input{days/tuesday_6_11_19.tex}



% ------------
% -- Tuesday --
% ------------
\newpage
\section{Wednesday June 12th: Main Conference}
\input{days/wednesday_6_12_19.tex}



% --------------
% -- Thursday --
% --------------
\newpage
\section{Thursday June 13th: Main Conference}
\input{days/thursday_6_13_19.tex}



% ------------
% -- Friday --
% ------------
\newpage
\section{Friday June 14th: Workshops}
\input{days/friday_6_14_19.tex}



\section*{Edits}

Many thanks to the following individuals for sending along helpful edits/catching typos:
\begin{itemize}
    \item Zeno Zantner for catching typos.
    \item Brandon Amos for catching typos.
\end{itemize}


% --- Bibliography ---
\newpage
\bibliographystyle{plainnat}
\bibliography{icml}

\end{document}