thesis.tex

%%
%% GMU LaTeX MS Thesis Format Template
%%
%% Developed by:
%%      Daniel O. Awduche and Christopher A. St. Jean
%%      Communications and Networking Lab
%%      Dept. of Electrical and Computer Engineering
%%
%% Notes on usage can be found in the accompanying USAGE_NOTES.txt file.
%%
%%**********************************************************************
%% Legal Notice:
%% This code is offered as-is without any warranty either
%% expressed or implied; without even the implied warranty of
%% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE!
%% User assumes all risk.
%% In no event shall any contributor to this code be liable for any damages
%% or losses, including, but not limited to, incidental, consequential, or
%% any other damages, resulting from the use or misuse of any information
%% contained here.
%%**********************************************************************
%%
%% $Id: GMU_thesis_template.tex,v 1.16 2007/05/02 02:20:11 Owner Exp $
%%

\documentclass[11 pt]{report}

%%  The file ``gmuthesis.sty''  is the GMU latex style file and
%%   should be placed in the same directory as your LaTeX files
\usepackage{gmuthesis}

%%
%% other packages that need to be loaded
%%
\usepackage[activate=true]{microtype} %don't hyphenate everywhere
\usepackage{graphicx}                    %   for imported graphics
\usepackage{rotating} %to rotate figs
\usepackage{amsmath}                     %%
\usepackage{amsfonts}                    %%  for AMS mathematics
\usepackage{amssymb}                     %%
\usepackage{amsthm}                      %%
\usepackage[normalem]{ulem}              %   a nice standard underline package
\usepackage[noadjust,verbose,sort]{cite} %   arranges reference citations neatly
\usepackage{setspace}                    %   for line spacing commands
\usepackage{float}                       %for precise placemnt of figs
\usepackage{subcaption} \usepackage{caption}
\usepackage{algorithmic} \usepackage{algorithm}
\usepackage{csvsimple}
\usepackage{hyperref}
\usepackage{enumitem}
% circled numbers for footnotes b/c it's confusing in math
\usepackage{pifont}
\renewcommand\thefootnote{\ding{\numexpr171+\value{footnote}}}
%for thick vertical line
\usepackage{array} \usepackage{makecell}
\usepackage[]{minted}
%http://tex.stackexchange.com/questions/237075/minted-not-working
%http://tex.stackexchange.com/questions/211672/cannot-get-tex-command-extra-options-to-work-in-auctex-11-88
\usepackage[table]{xcolor}

%% The file ``mythesisabbrev.sty'' is an (optional) personalized file that
%% may contain any and all LaTeX command (re)definitions that will be used
%% throughout the document
\usepackage{etoolbox}  \usepackage{mdframed}
\BeforeBeginEnvironment{minted}{\ls{1} \begin{mdframed}}%
\AfterEndEnvironment{minted}{\end{mdframed} \doublespacing}% ugly


\input{custom}

% % fig stuff does not work!
% \usepackage{standalone}
% \usepackage{pgf}
% \usepackage{tikz}

%so that i can have a standlone cv
\usepackage{standalone}
\usepackage{pdfpages}

\beforedoc

\usepackage{caption}
\captionsetup[subfigure]{labelformat=simple}
\renewcommand{\thesubfigure}{(\alph{subfigure})}
\renewcommand{\subfigurename}{Figure \thefigure}

\begin{document}

%% In this section, all of the user-specific fields to be used in the
%% title pages are set
\title{%
  Unsupervised Anomaly Detection in Sequences\\
  Using Long Short Term Memory \\
  Recurrent Neural Networks} %todo ok title? recurrent is redundant!
\onelinetitle{%
  Unsupervised Anomaly Detection in Sequences
  Using Long Short Term Memory 
  Recurrent Neural Networks
} %todo ok title?
\author{%
Majid S. alDosari
}
\degree{Master of Science}
\subject{Computational Science}
\doctype{Thesis}
\dept{Computational and Data Sciences}
\degreeyear{2016}


\firstdeg{Bachelor of Science}
\firstdegschool{Vanderbilt University}
\firstdegyear{2003}
\seconddeg{Master of Science}
\seconddegschool{Vanderbilt University}
\seconddegyear{2012}


% Note: semester name should be written in its full-form. For example, Fall Semester, not just Fall.
\degreesemester{Spring Semester}

\advisor{Dr. Kirk D. Borne} %todo: middle inital?
\firstmember{Dr. Estela Blaisten-Barojas}
\secondmember{Dr. Igor Griva} 
\depthead{Dr. Kevin Curtin (acting)}
\assocdean{Dr. Donna Fox}
\dean{Dr. Peggy Agouris}


%%
%% Introductory pages
%%

% Note: The signature sheet is set according to the requirements of the Volgenau School of
% Information Technology and Engineering. If your college/school requirement is different,
% please make appropriate changes in the "signaturepage" section of gmudissertation.sty file.
\signaturepage



\titlepage

% copyright technically optional but should be included in to avoid potential pagination problems
\copyrightpage

%%
%% Dedication page
%%

\dedicationpage

\noindent I dedicate this thesis to my father, Saad F. al-Dosari, who supported me in this endeavor.

%%
%% Acknowledgements
%%

\acknowledgementspage

\noindent
I appreciate Dr. Borne's lead, as well as encouraging enthusiasm, in this endeavor.
%
Similarly, I am grateful to my other committee members, Dr. Blaisten-Barojas and Dr. Griva, who have given their time and input so that I may successfully complete my thesis.
%
Also, I give special thanks to John Kaufhold of Deep Learning Analytics for being responsive to my questions regarding anything related to neural networks.
%
His interest in and support of my work motivated me to do the best job that I can.
%
Last but not least, I appreciate very much that Leif Johnson, author of the \textsf{Theanets} neural network package that I used, provided assistance beyond user support.


%%
%% Table of contents, list of tables, and lists of figures
%%

\tableofcontents

\listoftables

\listoffigures



%%
%% Abstract
%%
\abstractpage

Long Short Term Memory (LSTM) recurrent neural networks (RNNs) are evaluated for their potential to generically detect anomalies in sequences.
%
First, anomaly detection techniques are surveyed at a high level so that their shortcomings are exposed.
%
The shortcomings are mainly their inflexibility in the use of a context `window' size and/or their suboptimal performance in handling sequences.
%
Furthermore, high-performing techniques for sequences are usually associated with their respective knowledge domains.
%
After discussing these shortcomings, RNNs are exposed mathematically as generic sequence modelers that can handle sequences of arbitrary length.
%
From there, results from experiments using RNNs show their ability to detect anomalies in a set of test sequences.
%
The test sequences had different types of anomalies and unique normal behavior.
%
Given the characteristics of the test data, it was concluded that the RNNs were not only able to generically distinguish rare values in the data (out of context) but were also able to generically distinguish abnormal \emph{patterns} (in context).

\abstractmultiplepage
In addition to the anomaly detection work, a solution for reproducing computational research is described.
%
The solution addresses reproducing compute applications based on \textsf{Docker} container technology as well as automating the infrastructure that runs the applications.
%
By design, the solution allows the researcher to seamlessly transition from local (test) application execution to remote (production) execution because little distinction is made between local and remote execution.
%
Such flexibility and automation allows the researcher to be more confident of results and more productive, especially when dealing with multiple machines.

%% Be sure to leave a line of whitespace immediately before this line!!!!!
%% (If this comment segment runs together with the preceeding text, you might
%%  see the second page of the abstract numbered "0".)
%%
%% If the abstract is more than one page, then place this line PRECISELY
%% at the page break; otherwise, comment it out.  (See note about this line
%% in the usage notes.)
%%
%\abstractmultiplepage
%The second page of the abstract
%%
%%  the main body of the dissertation
%%
\startofchapters
%% include the chapters one by one (or paste the chapter text in directly if desired)
\include{introduction}
\include{ad}
\include{rnn}
\include{experiment}
\include{conclusions}
%%
%%  bibliography

%% list all of the BibTeX files here for the WinEdt project (if applicable)
%GATHER{bibfile.bib}

%% any bibliography style can be used, but IEEEtran.bst is ideally suited to
%% electrical engineering references





%% include the following directives if there are any appendices
\appendix
\appendixeqnumbering
\include{appendix/appendix}

\bibliographystyle{IEEEtran}
\bibliography{IEEEfull,library}

%%
%% curriculum vitae
%%
\cvpage
\input{cv}

\end{document}

%%% Local Variables:
%%% mode: latex
%%% TeX-master: t
%%% TeX-command-extra-options: "-shell-escape"
%%% End: