main.tex

%\documentclass[hidelinks,11pt,a4paper]{report}
\documentclass[PhD]{iiitd}
% \usepackage[left=3.0 cm,right=4.0 cm,top=4.0 cm,bottom=5.0 cm]{geometry}
\usepackage{svg} % this has to come above graphicx line
\usepackage[utf8]{inputenc}
\usepackage{lipsum}
\usepackage{xurl}
\usepackage{graphicx}
\usepackage{pdflscape}  % For rotating content
\usepackage{graphicx}   % Required for rotating tables
\usepackage[breaklinks=true]{hyperref}
%\usepackage{devanagari}
\usepackage[T1]{fontenc} % this is specially added to handle angel brackets
\usepackage{amsfonts} % for Z symbol
\usepackage{colortbl} % this is for colored columns
\usepackage{multirow} % for merging multiple rows in a table
\usepackage{multicol} % for merging multiple cols in a table
\usepackage{makecell} % for line breaks in table cell
\usepackage{booktabs} % this is for toprule and bottomrule in tables
\usepackage{xcolor} % color only one cell of a table
\usepackage{pdflscape} % one page landscape
%\usepackage[numbers,sort&compress]{natbib}
\usepackage{cite}
\usepackage{amsmath,amsfonts,amssymb}
\let\oldcite\cite
\renewcommand{\cite}[1]{\citep{#1}}
\input{math_commands.tex}
\setcounter{tocdepth}{5}
\setcounter{secnumdepth}{6}
\usepackage{hyperref}

\newcommand{\mathbbm}[1]{\text{\usefont{U}{bbm}{m}{n}#1}}
\input{behavior_model_macros}

\definecolor{gr}{RGB}{ 217, 234, 211 }
\definecolor{bl}{RGB}{ 201, 218, 248 }
\definecolor{rd}{RGB}{ 244, 204, 204 }
\definecolor{or}{RGB}{ 249, 203, 156 }


%%% For the xelatex (and other LaTeX friends) logos
\usepackage{hologo}

%%% For the awesome fontawesome icons!
\usepackage{fontawesome}


%%% For accessing system, OTF and TTF fonts
%%% (would have been loaded by polylossia anyway)
\usepackage{fontspec}
\usepackage{xunicode} %% loading this first to avoid clash with bidi/arabic

%%% For language switching -- like babel, but for xelatex
\usepackage{polyglossia}
\usepackage{placeins}
\setmainlanguage{english}
\setotherlanguages{hindi,sanskrit} %% or other languages
\newfontfamily\devanagarifont[Script=Devanagari]{Noto Serif Devanagari}
\usepackage{tabularx}
\usepackage{enumitem}

% Set global spacing for all enumerate environments
\setlist[enumerate]{topsep=0.5em} % Adjust the value of topset as needed



\begin{document}

% \maketitle
\input{cover}

\pagenumbering{roman}
\setcounter{page}{1}

% \chapter*{Certificate}
% \addcontentsline{toc}{chapter}{Certificate} 
% TEXT TO BE ADDED


\chapter*{Preface}
\addcontentsline{toc}{chapter}{Preface} 
\begin{table}[!th]
    \centering
    \begin{tabularx}{1\textwidth}{X}
        
        \textit{We're actually much better at planning the flight path of an interplanetary rocket (rocket science) than we are at managing the economy, merging two corporations, or even predicting how many copies of a book will sell (behavior prediction). So why is it that rocket science \textbf{seems} hard, whereas problems having to do with people - which arguably are much harder - seem like they ought to be \textbf{just} a matter of common sense (easily predictable)?} - Duncan J. Watts\\

        \begin{center}
                Also,
        \end{center}\\


        \textit{If the brain were so simple we could understand (predict) it, we would be so simple we couldn't.} - Emerson Pugh\\
        
        
        
        \begin{center}
            But,
        \end{center}\\
        
        \textit{Nothing in Nature is random (unpredictable). A thing appears random only through the incompleteness of our knowledge (ignorance).} - Baruch Spinoza\\

        \begin{center}
                While,
        \end{center}\\
        

        Ignorance is bliss. - Thomas Gray\\
        
        \begin{center}
                but,
        \end{center}\\

        \textit{Timendi causa est nescire.} (Ignorance is (also) the cause of fear.) - Seneca\\

        \begin{center}
                And,
        \end{center}\\
        
        \textit{What would life be if we had (only fear and) no courage to attempt anything?} - Vincent Van Gogh\\
        
    \end{tabularx}
    
\end{table}


As a computer scientist working on problems related to human behavior, I am often asked why I chose this particular domain. The questions come from various perspectives - whether such problems are better suited for psychologists and marketers, why these problems are interesting at all, whether human behavior contains too much randomness to be mathematically tractable, if the problems are ill-defined, and if they are even objectively solvable. For the ones interested in the art of diction, I try to lay out my motivations in the quotes above. But for others, I try to explain through a simple narrative and answer the questions more scientifically in Chapter-1. 



\textit{Behavior is unsolved}. Let me tell you a little story. Kelin is an eager advertiser who releases a campaign on Facebook one Friday evening, paying \$1000 to run an ad across California. With each launch, she silently sends a prayer that the ad resonates, draws clicks, leads to purchases of her product, ensures her campaign's success, and lands her the long-awaited promotion. Come Monday morning, Kelin witnesses human behavior in all its varied glory (within the platform's constraints, obviously): she has received 28 comments on her post, 867 likes, 9045 views, 349 clicks, and 28 purchases. Satisfied but seeking improvement, she tweaks a few words she feels might better appeal to Californians and relaunches. This time, her metrics jump by 10.8\%. Puzzled by the reasons, but pleased with the outcome, she presses on.


From my perspective, Kelin and countless others like her are replicating what the pioneering botanist Gregor Mendel did in the 1800s. The difference is that the subjects for Mendel were peas and for Kelin, it is humans. Mendel was trying to solve the puzzle of why some pea plants are tall and some small, some green while some yellow, and some pea seeds round while some wrinkled. The modern-day Kelins are trying to solve what makes people click, comment, like, and purchase, why certain words perform better in California while others in Texas, how behavior can be modified, and so on. Mendel's laboratory was his 2-acre Moravian monastery farm. Kelin's laboratory is the digital landscape of Facebook, Twitter, YouTube, TikTok, Google, and her website. 

Before Mendel, the general understanding of heredity was one of: \textit{Spontaneous Generation} (organisms could arise spontaneously from non-living matter), \textit{Lamarckism} (traits acquired by an organism during its lifetime could be passed on to its offspring), \textit{Blending Inheritance} (traits of offspring were a blend of the traits of their parents), and \textit{Preformationism} (miniature versions of organisms existed within the reproductive cells of parents). Today, 150 years later, we know to a very high degree of certainty, how traits in organisms arise and their mechanism of inheritance, to the point that we can calculate the probability of a certain type of rare cancer in the offspring of two given parents. However, Kelin's problem of who will click on her ad and how to maximize it remains unsolved and is often considered not worthy enough to be solved by science. Before the heroics of Mendel and Darwin, even the \textit{science} of heredity was considered a domain of philosophy and not worth the \textit{seriousness} of science. 


Rocket science is considered the hardest of sciences. \textit{It is solved.} It is solved to the extent that interplanetary launches over millions of kilometers can be planned to an accuracy of a few meters. Yet human conduct stays inscrutable, quirky, maddingly difficult to forecast and optimize for. It is unsolved to the extent that even opinion polls conducted right before the day of the election give opposite results to what is the actual outcome the next day. In my opinion, if behavior is not the problem to be worth solving, then what is!




%The world is its own best model. - Rodney Brooks

\clearpage

%\chapter*{Abstract}
%\addcontentsline{toc}{chapter}{Abstract}  
%XXX: To be done
%\clearpage


% \chapter{Course Work}
% CSE660A - Trustworthy AI Systems [Credits - 2]
% Final Grade: A-

% CSE641 - Deep Learning [Credits - 4]
% Final Grade: B

% CSE556 - Natural Language Processing [Credits - 4]
% Final Grade: B

% CSE542 - Statistical Machine Learning [Credits - 4]
% Final Grade: A-

% PIS790A - Independent Study [Credits - 2]
% Final Grade: A

% CGPA: 8.63 




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

\chapter*{Acknowledgments}
\addcontentsline{toc}{chapter}{Acknowledgments}  
\begin{table}[th]
    \centering
    \begin{tabular}{c}
\begin{sanskrit}कार्पण्यदोषोपहतस्वभाव:\end{sanskrit}\\
\begin{sanskrit}पृच्छामि त्वां धर्मसम्मूढचेता: |\end{sanskrit}\\
\begin{sanskrit}यच्छ्रेय: स्यान्निश्चितं ब्रूहि तन्मे\end{sanskrit}\\
\begin{sanskrit}शिष्यस्तेऽहं शाधि मां त्वां प्रपन्नम् ||\end{sanskrit}BG \begin{sanskrit}2:7||\end{sanskrit}\\\\
\begin{sanskrit}मयि सर्वाणि कर्माणि संन्यस्याध्यात्मचेतसा |\end{sanskrit}\\
\begin{sanskrit}निराशीर्निर्ममो भूत्वा युध्यस्व विगतज्वर: ||\end{sanskrit}BG \begin{sanskrit}3:30||\end{sanskrit}\\\\
\begin{sanskrit}
    नैव किञ्चित्करोमीति युक्तो मन्येत तत्त्ववित् |\end{sanskrit}\\
\begin{sanskrit} पश्यञ्शृण्वन्स्पृशञ्जिघ्रन्नश्नन्गच्छन्स्वपञ्श्वसन् ||\end{sanskrit}\\
\begin{sanskrit}प्रलपन्विसृजन्गृह्ण्न्नुन्मिषन्निमिषन्नपि |\end{sanskrit}\\
\begin{sanskrit}इन्द्रियाणीन्द्रियार्थेषु वर्तन्त इति धारयन् ||\end{sanskrit}BG \begin{sanskrit}5:8-9||\end{sanskrit}\\
    
\end{tabular}
\end{table}


This is an attempt to capture and thank those who have shaped my journey. Albeit, due to indirect and latent relations, this list will remain non-exhaustive despite my best attempts, it is still an attempt worth making.

In no particular order (with names of the organizations where I met these giants): Rajiv Ratn Shah (IIIT-D), Changyou Chen (SUNY at Buffalo), Ranjeeta Rani (GMPS), Roger Zimmerman (National University of Singapore), Debanjan Mahata (Bloomberg), Jessy Junyi Li (University of Texas at Austin), Balaji Krishnamurthy (Adobe MDSR), Sridhar Gantimahapatruni (Adobe), Jayakumar Subramanian (Adobe MDSR), Amanda Stent (Bloomberg), Anil Seth (FIITJEE), Dhruva Sahrawat (IIIT-D), Yifang Yin (National University of Singapore), Mika Hama (Second Language Testing Institute), Payman Vafaee (Columbia University), Pankaj Bansal (Adobe), Mohit Srivastava (Adobe), Gaurav Jain (Adobe), Shubham Yadav (NSIT), Rohit Jain (NSIT), Mohd Khwaja Salik (NSIT), Pratham Nawal (NSIT), Mayank Singh (NSIT), Somesh Singh (BITS-Pilani Goa), Aanisha Bhattacharyya (Adobe MDSR), Varun Khurana (IIIT-D), Rita Yadav (GMPS), Prabha Sinha (GMPS), Neeta Pandit (GMPS), Geetha Nair (GMPS), Swami Sarvapriyananda (Ramakrishna Mission), and finally my parents, Sushil Singla and Neena Singla, and my brother, Aman Singla.


Hopefully, I can return whatever I have gathered from these individuals back to society.
\clearpage


\include{abstract}
\clearpage

{\small
\tableofcontents
\addcontentsline{toc}{chapter}{Detailed Table of Contents}
}
\pagenumbering{gobble}


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\include{introduction}
\include{chapter-explaining-behavior}
\include{chapter-content-behavior-model}
\include{chapter-encoding-behavior-to-improve-content-understanding}
\include{chapter-generating-content-optimize-behavior}
\include{Conclusion}
\include{publications}






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

\listoftables
\addcontentsline{toc}{chapter}{List of Tables}
\listoffigures
\addcontentsline{toc}{chapter}{List of Figures}
%\listofsymbols
%\addcontentsline{toc}{chapter}{LIST OF SYMBOLS}
\begin{singlespace}
\bibliography{ref}
\addcontentsline{toc}{chapter}{Bibliography}
%\bibliographystyle{plainnat}
\end{singlespace}

\end{document}