main.tex

\documentclass{report}
\usepackage{setspace}
%\usepackage{subfigure}

\pagestyle{plain}
\usepackage{amssymb,graphicx,color}
\usepackage{amsfonts}
\usepackage{latexsym}
\usepackage{a4wide}
\usepackage{amsmath}
\usepackage{cancel}
\usepackage{tcolorbox}
\usepackage{caption}
\usepackage{float}
\usepackage{changepage}
\usepackage{hyperref}
\usepackage[perpage]{footmisc}
\usepackage{lipsum} % package that provides filler text
\usepackage{subcaption}
\usepackage{algorithm} 
\usepackage{algpseudocode} 


\tcolorboxenvironment{definition}{
  sharp corners,
  boxrule=0.4pt,
  colback=white,
  before skip=\topsep,
  after skip=\topsep,
}

\usepackage{tikz}
\usetikzlibrary{bayesnet}
\usetikzlibrary{arrows}
\usetikzlibrary{calc}
\usetikzlibrary{shadows}
\usetikzlibrary{positioning}

\usepackage{prerex}
\usetikzlibrary{fit}

\usepackage[english]{babel}
\usepackage{blindtext}
\usepackage{epigraph}


\newtheorem{theorem}{Theorem}
\newtheorem{lemma}[theorem]{Lemma}
\newtheorem{corollary}[theorem]{Corollary}
\newtheorem{proposition}[theorem]{Proposition}
\newtheorem{remark}[theorem]{Remark}
\newtheorem{definition}[theorem]{Definition}
\newtheorem{fact}[theorem]{Fact}

\newtheorem{problem}[theorem]{PROBLEM}
\newtheorem{exercise}[theorem]{EXERCISE}
\def \set#1{\{#1\} }

\newenvironment{proof}{
PROOF:
\begin{quotation}}{
$\Box$ \end{quotation}}

\newcommand{\Phu}[1]{{\bf \color{red} [[Phu: #1]]}}
% \newcommand{\Phu}[1]{{}}


\newcommand{\dby}{\ \mathrm{d}}
\newcommand{\kl}{D_{\mathrm{KL}}}
\newcommand{\argmax}[1]{\underset{#1}{\arg\max \ }}
\newcommand{\argmin}[1]{\underset{#1}{\arg\min \ }}
\newcommand{\const}{\text{const.}}
\newcommand{\optim}{\mathcal{O}}
\newcommand{\prior}{\text{prior}}
\newcommand{\bracka}[1]{\left( #1 \right)}
\newcommand{\brackb}[1]{\left[ #1 \right]}
\newcommand{\brackc}[1]{\left\{ #1 \right\}}
\newcommand{\contractop}{\mathcal{B}}
\newcommand*\circled[1]{\tikz[baseline=(char.base)]{
            \node[shape=circle,draw,inner sep=2pt] (char) {#1};}}
\newcommand{\red}[1]{{\color{red} #1}}
\newcommand{\loss}{\mathcal{L}}
\newcommand{\pri}{\operatorname{pri}}
\newcommand{\pub}{\operatorname{pub}}
\newcommand{\correctquote}[1]{``#1''}
\newcommand{\mseloss}[3]{\mathbb{E}_{#1}\brackb{\frac{1}{2}\bracka{#2 - #3}^2}}
\newcommand{\klloss}[2]{\kl\bracka{#1 \Bigg\| #2 }}
\newcommand{\norm}[1]{\left\lVert#1\right\rVert}

\newcommand{\grammar}[1]{{\color{blue} #1}}
\newcommand{\factcheck}[1]{{\color{magenta} #1}}

\newcommand{\nats}{\mbox{\( \mathbb N \)}}
\newcommand{\rat}{\mbox{\(\mathbb Q\)}}
\newcommand{\rats}{\mbox{\(\mathbb Q\)}}
\newcommand{\reals}{\mbox{\(\mathbb R\)}}
\newcommand{\ints}{\mbox{\(\mathbb Z\)}}

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


\title{{{\Huge Multi-agent Learning with Bayesian Inference}}\\
% {\large Optional Subtitle}\\
}
\date{Submission date: 18 May 2020}
\author{Phu Sakulwongtana\thanks{
{\bf Disclaimer:}
This report is submitted as part requirement for the BSc Computer Science at UCL. It is
substantially the result of my own work except where explicitly indicated in the text.
% \emph{Either:} 
The report may be freely copied and distributed provided the source is explicitly acknowledged
% \newline  %% \\ messes it up
% \emph{Or:}\newline
% The report will be distributed to the internal and external examiners, but thereafter may not be copied or distributed except with permission from the author.
}
\\ \\
BSc Computer Science\\ \\
Jun Wang, Philip Treleaven}

\newenvironment{outline} {\renewcommand\abstractname{}\begin{abstract}} {\end{abstract}}

\newenvironment{miniabstract} {\begin{adjustwidth}{1cm}{}} {\end{adjustwidth}}

\usepackage[includefoot,bottom=20pt]{geometry}
\addtolength{\skip\footins}{2pc plus 3pt}
\makeatletter
\renewenvironment{abstract}{%
    \if@twocolumn
      \section*{\abstractname}%
    \else %% <- here I've removed \small
      \begin{center}%
        {\bfseries \Large\abstractname\vspace{\z@}}%  %% <- here I've added \Large
      \end{center}%
      \quotation
    \fi}
    {\if@twocolumn\else\endquotation\fi}
\makeatother

\begin{document}
 
\onehalfspacing
\maketitle

\begin{abstract}
\input{non_content/abstract}
\end{abstract}

% \clearpage
% \vspace*{\fill}
% \thispagestyle{empty} % optional -- suppress showing of page number
% \begin{quotation}
% % \em % optional -- to switch to emphasis (italics) mode
% \correctquote{but what analogy to do you detect in the inquiry about justice?} \correctquote{I will tell you,} I said: \correctquote{there is a justice of one man, we say, and, I suppose, also of an entire city.} \correctquote{Assuredly,} said he. \correctquote{Is not the city larger than the man?} \correctquote{It is larger,} he said. \correctquote{Then, perhaps, there would be more justice in the larger object and more easy to apprehend. If it please you, then, let us first look for its quality in states, and then only examine it also in the individual, looking for the likeness of the greater in the form of the less.} \correctquote{I think that is a good suggestion,} he said. \correctquote{If, then,} said I, \correctquote{our argument should observe the origin of a state, we should see also the origin of justice and injustice in it.} \correctquote{It may be,} said he. \correctquote{And if this is done, we may expect to find more easily what we are seeking?}

% \medskip
% \raggedleft
% Plato, Republic 368e - 369a \\
% Translated by Paul Shorey \cite{gregory_crane}
% \end{quotation}
% \vspace*{\fill}

\setcounter{tocdepth}{1}
\tableofcontents

% \begin{outline}
% \input{non_content/chapter_outline}
% \end{outline}

\addtocontents{toc}{\protect\thispagestyle{empty}}

\setcounter{page}{1}

\chapter{Introduction}
\input{chapters/chapter01/main}

\chapter{Background}
\input{chapters/chapter02/main}

\chapter{Unified View of Probabilistic Multi-agent Reinforcement Learning}
\input{chapters/chapter03/main}

\chapter{Hierarchy and Communication in Multi-Agent Reinforcement Learning}
\input{chapters/chapter04/main}

\chapter{Probabilistic perspective on Public Belief MDP}
\input{chapters/chapter05/main}

\chapter{Conclusion and Future Work}
\input{chapters/chapter06/main}

\appendix

% \bibliographystyle{apalike}
\bibliographystyle{plain}
\bibliography{bibilography}

\chapter{Chapter 2: Appendix}

\section {Probabilistic Machine Learning}
\input{chapters/chapter02/appendix/01_proof}

\section {Single Agent Reinforcement Learning}
\input{chapters/chapter02/appendix/02_proof}

\section {Control as Inference Framework}
\input{chapters/chapter02/appendix/04_proof}

\chapter{Chapter 3: Appendix}

\section {ROMMEO}
\input{chapters/chapter03/appendix/02_proof}

\section {Balancing-Q}
\input{chapters/chapter03/appendix/04_proof}

\section{Experimentation}
\input{chapters/chapter03/appendix/06_proof}

\chapter{Chapter 4: Appendix}
\input{chapters/chapter04/appendix/01_proof}
\input{chapters/chapter04/appendix/02_proof}


% \chapter{Reinforcement Learning}
% \input{appendix/appendix_2.tex}

\end{document}