machine_learning.tex

\documentclass[a4paper]{book}
\usepackage{amsmath, amssymb, amsbsy}
\usepackage[amsmath, thmmarks]{ntheorem}
\usepackage{algorithm}
\usepackage{algpseudocode}
\usepackage{xcolor}
\usepackage{tikz}
\usepackage{paralist}
\usepackage{nicefrac}
\usepackage{minted}
\usepackage[%
    bookmarksnumbered,%
    bookmarksopen,%
    pdftitle={Machine Learning Notes},%
    pdfauthor=DeepWalter,%
    colorlinks=true,%
    pagebackref=true,
    hyperfootnotes=false % no hyperlink for footnotes.
]{hyperref}
\usepackage[adjust, nocompress, space]{cite}
\usepackage{setspace}
\usepackage[perpage, flushmargin, symbol*]{footmisc}
\usepackage{array}
\usepackage{diagbox}
\usepackage{booktabs}




% ============================================================================================================
% ============================================================================================================

% ============================= Spacing ========================================
% ========== Row Spacing ============
% Packages: setspace
\setstretch{1.2} % with extra 20% row spacing.


% ============================= Pseudocode Customization =======================
% Packages: algorithm, algpseudocode.
\algrenewcommand{\algorithmicrequire}{\textbf{Input:}}
\algrenewcommand{\algorithmicensure}{\textbf{Output:}}
\algnewcommand\algorithmicto{\textbf{to} } % a trailing space is needed.
\algnewcommand\algorithmicbreak{\textbf{break}} % break keyword.
\newcommand{\breakif}[1]{\State \textbf{if} {#1} \textbf{then break}}
\newcommand{\continueif}[1]{\State \textbf{if} {#1} \textbf{then continue}}


% ============================= Math Theorem Envs ==============================
\theorembodyfont{\normalfont}
\newtheorem{df}{Definition}[chapter]
\newtheorem{thm}[df]{Theorem}
\newtheorem{eg}[df]{Example}
\newtheorem{prop}[df]{Proposition}
\newtheorem{lem}[df]{Lemma}
\newtheorem{cor}[df]{Corollary}
\newtheorem{re}[df]{Remark}

% Customize proof env.
\theoremstyle{nonumberplain} % no numbering
\theoremsymbol{$\square$}    % proof end with square.
\newtheorem{pf}{Proof}


% =============================== Operators ====================================
\DeclareMathOperator{\ent}{Entropy}
\DeclareMathOperator{\gain}{Gain}
\DeclareMathOperator*{\p}{\mathbb{P}} % probability.
\DeclareMathOperator{\sign}{sign}
\DeclareMathOperator*{\argmax}{argmax}
\DeclareMathOperator*{\argmin}{argmin}
\DeclareMathOperator*{\E}{\mathbb{E}} % Expectation.
\DeclareMathOperator{\dist}{dist} % distance.
\DeclareMathOperator{\indi}{\mathbb{I}}
\DeclareMathOperator{\tr}{tr} % trace of an matrix.



% =============================== New Commands =================================
% Logical connectives:
\newcommand{\OR}{\textbf{OR}\ } % a space is needed here.
\newcommand{\AND}{\textbf{AND}\ }
\newcommand{\XOR}{\textbf{XOR}\ }
\newcommand{\NOT}{\textbf{NOT}\ }


% Math commands:
\newcommand{\T}[1]{\ensuremath{{#1}^\mathsf{T}}} % Matrix transposition.
\newcommand{\inv}[1]{\ensuremath{{#1}^{-1}}} % Inversion.
\newcommand{\V}[1]{\ensuremath{\boldsymbol{#1}}}
\newcommand{\forany}{\ensuremath{\forall~}} % append a non-breakable whitespace.
\newcommand{\thereis}{\ensuremath{\exists~}}
\newcommand{\linspan}[1]{\ensuremath{\operatorname{\mathbf{span}}\left\{{#1}\right\}}}
\newcommand{\diff}[1]{\ensuremath{\mathrm{d}{#1}}}
\newcommand{\norm}[2][2]{\ensuremath{||{#2}||_{#1}}}

% Shorthand commands:
\newcommand{\hypo}[1]{\ensuremath{{#1}: \mathcal{X} \longrightarrow \mathcal{Y}}} % Hypothesis function.
\newcommand{\dataset}{\ensuremath{D = \{(\V{x}_1, y_1), \dotsc, (\V{x}_m, y_m)\}}} % Labeled training set.
\newcommand{\st}{\text{s.t.\ }}  % such that
\newcommand{\wrt}{\text{w.r.t.\ }}  % with respect to
\newcommand{\iid}{\text{i.i.d.\ }}  % identically and independently distributed
\newcommand{\magenta}[1]{\textcolor{magenta}{#1}} % turn the color of text into magenta.
\newcommand{\hl}[2][yellow]{\colorbox{#1}{#2}} % highlight the text.
\newcommand{\pfrac}[2]{\ensuremath{\frac{\partial {#1}}{\partial {#2}}}}
\newcommand{\angpair}[2]{\ensuremath{\langle {#1}, {#2}\rangle}} % inner product with angles.
\newcommand{\marginnote}[1]{\marginpar{\scriptsize\itshape {#1}}}


% ============================= Python Code Customization ======================
% Package: minted
\setminted[python]{frame=single, linenos=true, numbersep=3pt, autogobble=true, mathescape=true}


% ========================================================================================================
% ============================================= Document =================================================
% ========================================================================================================
\begin{document}

% =====================Title Page====================
\title{\textbf{Machine Learning Notes}}
\author{DeepWalter\thanks{Email: deepwalter.cn@gmail.com}}
\date{}
\maketitle

\frontmatter
% =====================Table of Contents=============
\tableofcontents

% ===================================================
\mainmatter
% =============================Part One: Basics========================
\part{Basics}
% ========Model Assessment and Selection======
\chapter{Model Assessment and Selection}
\include{part_1/assessment/model_assessment_selection}

% ========Linear Models================
\chapter{Linear Model}
\include{part_1/linear/linear_model}

% ========Decision Trees===============
\chapter{Decision Tree}
\include{part_1/tree/decision_tree}

% ========SVM========================
\chapter{Support Vector Machine}
\include{part_1/svm/svm}

% ========Neuron Network
\chapter{Neuron Networks}
\include{part_1/nn/fc_neuron_network}
\include{part_1/nn/cnn}

% ========Bayesian Classifier=========
\chapter{Bayesian}
\include{part_1/bayes/bayesian_classifier}

% ========Clustering=================
\chapter{Clustering}
\include{part_1/clustering/clustering}

% ========Expectation Maximalization====
\chapter{Expectation Maximization}
\include{part_1/em/EM}

% ========Ensemble Learning============
\chapter{Ensemble Learning}
\include{part_1/ensemble/ensemble_learning}

% ========Dimension Reduction=========
\chapter{Dimension Reduction}
\include{part_1/reduction/dimension_reduction}

% ========Computational Learning Theory
\chapter{Computational Learning Theory}
\include{part_1/clt/CLT}

% ========Reinforcement Learning
\chapter{Reinforcement Learning}
\include{part_1/reinforcement/reinforcement_learning}

% =============================Part Two: Papers====================
\part{Selected Papers}

\chapter{Neuron Networks}
\include{part_2/nn/hornik_1989}
\include{part_2/nn/cybenko_1989}

%============================================================
\backmatter
% ========================Bibliography=======================
\addcontentsline{toc}{part}{Bibliography}
\bibliographystyle{amsplain}
\bibliography{notes}


\end{document}