\relax 
\citation{me05a}
\citation{boehm00a}
\citation{andrews07}
\citation{holzmann97}
\citation{port08}
\citation{endres03}
\citation{endres03}
\citation{fenton07}
\citation{witten05}
\@writefile{toc}{\contentsline {chapter}{\numberline {1}Introduction}{1}}
\@writefile{lof}{\addvspace {10\p@ }}
\@writefile{lot}{\addvspace {10\p@ }}
\newlabel{introduction}{{1}{1}}
\@writefile{toc}{\contentsline {section}{\numberline {1.1}Motivation}{1}}
\citation{norvig03,She02,leake96,kol93,leake05}
\citation{me00f,me07g,me07f,orrego09,me09a,me09e,me09f,me09i,me10f}
\citation{ruhe09,zhang07a}
\citation{ruhe09}
\citation{shepperd97,Mendes2003,Kirsopp2002,Walkerden1999}
\citation{me00f,me07g,me07f,orrego09,me09a,me09e,me09f,me09i,me10f}
\citation{brady10a}
\citation{brady10b}
\@writefile{toc}{\contentsline {section}{\numberline {1.2}Statement of Thesis}{4}}
\@writefile{toc}{\contentsline {section}{\numberline {1.3}Contribution of This Work}{4}}
\@writefile{toc}{\contentsline {section}{\numberline {1.4}Document Structure}{5}}
\citation{shepperd97,Mendes2003,Kirsopp2002,Walkerden1999}
\citation{pendharkar05,Li2009,Lipowezky1998,Walkerden1999,Kirsopp2002,Mendes2003,jorgensen05,shepperd07,shepperd97}
\citation{pendharkar05}
\citation{fenton99}
\citation{harman04}
\citation{harman04}
\citation{hall03}
\citation{miller02}
\citation{Dil84}
\citation{hall03}
\@writefile{toc}{\contentsline {chapter}{\numberline {2}Related Work}{7}}
\@writefile{lof}{\addvspace {10\p@ }}
\@writefile{lot}{\addvspace {10\p@ }}
\newlabel{related}{{2}{7}}
\@writefile{toc}{\contentsline {section}{\numberline {2.1}Software Estimation Research}{7}}
\citation{jalali08}
\citation{andrews07}
\citation{me07f}
\citation{gof95}
\citation{pfahl05}
\citation{me02f}
\citation{boehm00b}
\@writefile{toc}{\contentsline {section}{\numberline {2.2}Search-Based Software Engineering (SBSE)}{8}}
\@writefile{toc}{\contentsline {section}{\numberline {2.3}Model: Benefits}{8}}
\citation{fenton07}
\citation{chulani99}
\citation{schulz10}
\citation{me07f}
\@writefile{toc}{\contentsline {section}{\numberline {2.4}Model: Drawbacks}{9}}
\newlabel{sect:draw}{{2.4}{9}}
\@writefile{lof}{\contentsline {figure}{\numberline {2.1}{\ignorespaces Features of the COCOMO model ontology.}}{10}}
\newlabel{fig:cocattrs}{{2.1}{10}}
\citation{lowry10}
\citation{baker07}
\citation{boehm00b}
\citation{baker07}
\citation{boehm81}
\newlabel{eq:simpcoc}{{2.1}{11}}
\newlabel{eq:ab}{{2.2}{11}}
\citation{me05a}
\citation{me05a}
\citation{me05a}
\citation{me05a}
\@writefile{lof}{\contentsline {figure}{\numberline {2.2}{\ignorespaces COCOMO 1 effort multipliers, and the sorted coefficients found by linear regression from twenty 66\% sub-samples (selected at random) from the NASA93 PROMISE data set; from\nobreakspace  {}\cite  {me05a}. }}{13}}
\newlabel{fig:coc1}{{2.2}{13}}
\citation{bartlett32}
\citation{loftus03}
\@writefile{toc}{\contentsline {chapter}{\numberline {3}The {$\mathcal  {W}$}2 Algorithm}{14}}
\@writefile{lof}{\addvspace {10\p@ }}
\@writefile{lot}{\addvspace {10\p@ }}
\newlabel{sect:inst}{{3}{14}}
\@writefile{toc}{\contentsline {section}{\numberline {3.1}Case-Based Reasoning}{14}}
\citation{kolodner83}
\citation{aamod94}
\@writefile{lof}{\contentsline {figure}{\numberline {3.1}{\ignorespaces Four steps of case-based reasoning, from \url {http://www.peerscience.com/intro_cbr.htm}.}}{15}}
\newlabel{fig:cbr}{{3.1}{15}}
\citation{schank83}
\@writefile{toc}{\contentsline {section}{\numberline {3.2}Contrast Set Learning (CSL)}{16}}
\newlabel{sect:csl-description}{{3.2}{16}}
\citation{me09f}
\citation{mozina04}
\citation{me10c}
\newlabel{eq:br}{{3.1}{18}}
\@writefile{toc}{\contentsline {section}{\numberline {3.3}The {$\mathcal  {W}$}2 Algorithm}{18}}
\newlabel{sect:w-desc}{{3.3}{18}}
\citation{brooks75}
\@writefile{lof}{\contentsline {figure}{\numberline {3.2}{\ignorespaces The Brooks' Law Query for the NASA93 dataset in COCOMO II format.}}{19}}
\newlabel{fig:brooks}{{3.2}{19}}
\@writefile{lof}{\contentsline {figure}{\numberline {3.3}{\ignorespaces RELEVANT= cases nearest to $context_1$.}}{20}}
\newlabel{fig:train-filter}{{3.3}{20}}
\@writefile{lof}{\contentsline {figure}{\numberline {3.4}{\ignorespaces $Best$ (top) \& $rest$ (bottom).}}{21}}
\newlabel{fig:train-contrast}{{3.4}{21}}
\@writefile{toc}{\contentsline {subsection}{\numberline {3.3.1}Relevancy Filtering}{21}}
\@writefile{toc}{\contentsline {subsection}{\numberline {3.3.2}Utility Separation}{21}}
\@writefile{lof}{\contentsline {figure}{\numberline {3.5}{\ignorespaces Rank with Equation\nobreakspace  {}3.1\hbox {}.}}{22}}
\newlabel{fig:train-bestrest}{{3.5}{22}}
\@writefile{toc}{\contentsline {subsection}{\numberline {3.3.3}Contrast Set Generation}{22}}
\@writefile{lof}{\contentsline {figure}{\numberline {3.6}{\ignorespaces A $K_1=20$ neighborhood of $context_1$ (NASA93ii train set).}}{23}}
\newlabel{fig:testing}{{3.6}{23}}
\@writefile{toc}{\contentsline {subsection}{\numberline {3.3.4}Estimating Impact}{23}}
\@writefile{lof}{\contentsline {figure}{\numberline {3.7}{\ignorespaces All rows of Figure\nobreakspace  {}3.6\hbox {} satisfying $R_1 : pmat=3$. }}{24}}
\newlabel{fig:testing-constrained}{{3.7}{24}}
\@writefile{lof}{\contentsline {figure}{\numberline {3.8}{\ignorespaces The testing set with all cases not containing $pmat = 3$ removed. }}{24}}
\newlabel{fig:testing-applied}{{3.8}{24}}
\@writefile{lof}{\contentsline {figure}{\numberline {3.9}{\ignorespaces Result of applying the learned constraint $pmat = 3$ to the Brooks' Law query $q$ during testing. The median estimate reduction from 235 to 81 represents a 66\% reduction is software effort by applying $pmat = 3$.}}{24}}
\newlabel{fig:testing-result}{{3.9}{24}}
\citation{shepperd97}
\citation{watson98}
\@writefile{lof}{\contentsline {figure}{\numberline {3.10}{\ignorespaces Revising $q$ to learn $q'$.}}{25}}
\newlabel{fig:w}{{3.10}{25}}
\@writefile{lof}{\contentsline {figure}{\numberline {3.11}{\ignorespaces {$\mathcal  {W}$}2's syntax for describing the input query $q$. Here, all the values run 1 to 6. $4\le cplx \le 6$ denotes projects with above average complexity. Question marks denote what can be controlled- in this case, $rely,time$ (required reliability and development time)}}{26}}
\newlabel{fig:nasaproj}{{3.11}{26}}
\@writefile{toc}{\contentsline {section}{\numberline {3.4}Measuring Performance}{26}}
\@writefile{toc}{\contentsline {chapter}{\numberline {4}Experiments with {$\mathcal  {W}$}2}{28}}
\@writefile{lof}{\addvspace {10\p@ }}
\@writefile{lot}{\addvspace {10\p@ }}
\newlabel{experiments}{{4}{28}}
\@writefile{toc}{\contentsline {section}{\numberline {4.1}Datasets and Project Descriptions}{28}}
\@writefile{lof}{\contentsline {figure}{\numberline {4.1}{\ignorespaces Seven data sets from \url {promisedata.org/?cat=14}: {\em  effort} is total staff person-months; {\em  time} is calendar time (start to stop); {\em  defects} represents the number of delivered defects. }}{28}}
\newlabel{fig:data}{{4.1}{28}}
\citation{miyazaki94}
\citation{kemerer87}
\citation{shepperd97}
\citation{brady10b}
\@writefile{lof}{\contentsline {figure}{\numberline {4.2}{\ignorespaces Example project $controllable$ file for Chinese software projects after discretization. Ranges were assigned randomly for this project. A ``?'' represents a controllable feature. If an attribute range isn't specified in the project, it is ignored.}}{30}}
\newlabel{fig:china-p2}{{4.2}{30}}
\@writefile{toc}{\contentsline {section}{\numberline {4.2}Experiment: {$\mathcal  {W}$}2 vs {$\mathcal  {W}$}}{30}}
\newlabel{sect:runtimes}{{4.2}{30}}
\@writefile{lof}{\contentsline {figure}{\numberline {4.3}{\ignorespaces Average execution times for the W and W2 algorithms. By removing the $O(n^2)$ kth nearest neighbor calculation from W we drastically improve performance, especially on larger datasets such as China (499 cases).}}{31}}
\newlabel{fig:w2-runtimes}{{4.3}{31}}
\@writefile{lof}{\contentsline {figure}{\numberline {4.4}{\ignorespaces Performance of W2's Overlap relevancy filtering vs W's kth nearest-neighbor filtering for 5 unique datasets.}}{31}}
\newlabel{fig:bestoverlap-vs-knn}{{4.4}{31}}
\@writefile{toc}{\contentsline {section}{\numberline {4.3}Experiment: {$\mathcal  {W}$}2's Performance Across Multiple Datasets}{32}}
\newlabel{sect:multi-comp}{{4.3}{32}}
\@writefile{lof}{\contentsline {figure}{\numberline {4.5}{\ignorespaces Effort estimation improvements ($100*\frac  {initial-final}{intial}$) for five unique datasets. Sorted by median improvement. Gray cells represent no improvement in effort estimates.}}{33}}
\newlabel{fig:other-reduce-values}{{4.5}{33}}
\@writefile{lof}{\contentsline {figure}{\numberline {4.6}{\ignorespaces Effort results for five non-COCOMO datasets. }}{34}}
\newlabel{fig:other-reduce}{{4.6}{34}}
\@writefile{toc}{\contentsline {section}{\numberline {4.4}Experiment: Intra- and Inter-Project Stability}{34}}
\newlabel{sect:stable}{{4.4}{34}}
\@writefile{lof}{\contentsline {figure}{\numberline {4.7}{\ignorespaces Range of changes in median and spread generated by applying the recommendations of {$\mathcal  {W}$}2. The median observed changes were (20.5, 20.5)\% for (medians, spreads), respectively. }}{35}}
\newlabel{fig:changes}{{4.7}{35}}
\citation{me10b}
\citation{zimmermann09}
\@writefile{lof}{\contentsline {figure}{\numberline {4.8}{\ignorespaces Recommendation frequency across 20 runs of {$\mathcal  {W}$}2 for reducing individual goals ($defects$, $effort$, or $months$) as well as all goals at once ($all$).}}{36}}
\newlabel{fig:stability}{{4.8}{36}}
\citation{me09a}
\citation{me09a}
\@writefile{toc}{\contentsline {section}{\numberline {4.5}Experiment: Comparing Drastic Changes to {$\mathcal  {W}$}2}{37}}
\newlabel{sect:drastic-comp}{{4.5}{37}}
\@writefile{lof}{\contentsline {figure}{\numberline {4.9}{\ignorespaces Examples of drastic changes to software projects.}}{38}}
\newlabel{fig:drastic-overview}{{4.9}{38}}
\@writefile{lof}{\contentsline {figure}{\numberline {4.10}{\ignorespaces Comparing defect, effort, and month estimation reduction percentages ($100*\frac  {initial-final}{intial}$ of drastic business decisions vs {$\mathcal  {W}$}'s recommendations for the Ground case study.}}{38}}
\newlabel{fig:drastic-ground}{{4.10}{38}}
\@writefile{lof}{\contentsline {figure}{\numberline {4.11}{\ignorespaces Comparing defect, effort, and month estimation reduction percentages ($100*\frac  {initial-final}{intial}$ of drastic business decisions vs {$\mathcal  {W}$}'s recommendations for the Flight case study.}}{39}}
\newlabel{fig:drastic-flight}{{4.11}{39}}
\@writefile{lof}{\contentsline {figure}{\numberline {4.12}{\ignorespaces Comparing defect, effort, and month estimation reduction percentages ($100*\frac  {initial-final}{intial}$ of drastic business decisions vs {$\mathcal  {W}$}'s recommendations for the OSP case study.}}{40}}
\newlabel{fig:drastic-osp}{{4.12}{40}}
\@writefile{lof}{\contentsline {figure}{\numberline {4.13}{\ignorespaces Comparing defect, effort, and month estimation reduction percentages ($100*\frac  {initial-final}{intial}$ of drastic business decisions vs {$\mathcal  {W}$}'s recommendations for the OSP2 case study.}}{41}}
\newlabel{fig:drastic-osp2}{{4.13}{41}}
\@writefile{toc}{\contentsline {chapter}{\numberline {5}Model-Based vs. Case-Based Algorithms}{42}}
\@writefile{lof}{\addvspace {10\p@ }}
\@writefile{lot}{\addvspace {10\p@ }}
\newlabel{sect:modelvinstance}{{5}{42}}
\@writefile{toc}{\contentsline {section}{\numberline {5.1}Model-based Case Studies}{42}}
\newlabel{sect:data}{{5.1}{42}}
\@writefile{lof}{\contentsline {figure}{\numberline {5.1}{\ignorespaces Contexts of 4 case studies. \textit  {\{1, 2, 3, 4, 5, 6\}} map to \textit  {\{very low, low, nominal, high, very high, extra high\}}.}}{43}}
\newlabel{fig:cases}{{5.1}{43}}
\citation{me07f}
\@writefile{toc}{\contentsline {section}{\numberline {5.2}SEESAW}{44}}
\newlabel{eq:pred}{{5.1}{44}}
\newlabel{eq:simp}{{5.2}{44}}
\citation{boehm00b}
\newlabel{eq:argmin}{{5.4}{45}}
\newlabel{eq:lineem}{{5.5}{45}}
\citation{boehm00b}
\citation{me07f,me08c,me09e}
\citation{me09a}
\citation{me09i}
\newlabel{eq:linesf}{{5.6}{46}}
\citation{me07f}
\citation{kirkpatrick83}
\@writefile{lof}{\contentsline {figure}{\numberline {5.2}{\ignorespaces Pseudocode for SEESAW}}{48}}
\newlabel{fig:seesaw}{{5.2}{48}}
\citation{selman96}
\@writefile{toc}{\contentsline {section}{\numberline {5.3}Five Additional AI Model-Based Algorithms}{49}}
\newlabel{eq:edt}{{5.7}{49}}
\citation{craw94}
\@writefile{lof}{\contentsline {figure}{\numberline {5.3}{\ignorespaces Example of SA's forward and back select.}}{50}}
\newlabel{fig:run}{{5.3}{50}}
\@writefile{toc}{\contentsline {section}{\numberline {5.4}Comparisons of AI Model-based Methods}{51}}
\newlabel{sect:aicomp}{{5.4}{51}}
\citation{boehm81}
\@writefile{lof}{\contentsline {figure}{\numberline {5.4}{\ignorespaces Number of times algorithms were top-ranked (largest is 4: i.e. one for each Figure\nobreakspace  {}5.1\hbox {} case study). }}{52}}
\newlabel{fig:xper}{{5.4}{52}}
\@writefile{toc}{\contentsline {section}{\numberline {5.5}Model vs. Case-Based Methods}{52}}
\newlabel{sect:modelins}{{5.5}{52}}
\@writefile{lof}{\contentsline {figure}{\numberline {5.5}{\ignorespaces Changes in median and spread for the NASA93 dataset. }}{54}}
\newlabel{fig:nasa93results}{{5.5}{54}}
\@writefile{lof}{\contentsline {figure}{\numberline {5.6}{\ignorespaces Changes in median and spread for the COC81 dataset. }}{55}}
\newlabel{fig:cocomo81results}{{5.6}{55}}
\citation{boehm00b}
\citation{boehm00a}
\@writefile{toc}{\contentsline {chapter}{\numberline {6}Discussion}{57}}
\@writefile{lof}{\addvspace {10\p@ }}
\@writefile{lot}{\addvspace {10\p@ }}
\newlabel{discussion}{{6}{57}}
\@writefile{toc}{\contentsline {section}{\numberline {6.1}When Not to Use {$\mathcal  {W}$}2}{57}}
\newlabel{sect:delta}{{6.1}{57}}
\citation{boehm00a}
\citation{Boehm96softwarecost}
\citation{shank77}
\citation{shepperd07}
\@writefile{lof}{\contentsline {figure}{\numberline {6.1}{\ignorespaces Relative effects on development effort. From\nobreakspace  {}\cite  {boehm00a}. }}{59}}
\newlabel{fig:coceffects}{{6.1}{59}}
\@writefile{toc}{\contentsline {section}{\numberline {6.2}Model-lite}{59}}
\newlabel{sect:modellite}{{6.2}{59}}
\citation{wolpert94}
\@writefile{toc}{\contentsline {section}{\numberline {6.3}Scope of the Study}{60}}
\citation{Li2009,Lipowezky1998,Walkerden1999,Kirsopp2002,Mendes2003}
\citation{ostrand04,journals/ese/KhoshgoftaarS03,tosun10,me10a}
\citation{bartlett32}
\citation{kolodner83}
\citation{shank77}
\citation{azzeh08}
\@writefile{toc}{\contentsline {chapter}{\numberline {7}Conclusion}{62}}
\@writefile{lof}{\addvspace {10\p@ }}
\@writefile{lot}{\addvspace {10\p@ }}
\newlabel{conclusion}{{7}{62}}
\citation{me10c}
\citation{me10c}
\citation{me01c}
\@writefile{toc}{\contentsline {chapter}{\numberline {A}{$\mathcal  {W}$}2 Source Code}{65}}
\@writefile{lof}{\addvspace {10\p@ }}
\@writefile{lot}{\addvspace {10\p@ }}
\newlabel{source}{{A}{65}}
\@writefile{toc}{\contentsline {section}{\numberline {A.1}w.sh}{65}}
\@writefile{lol}{\contentsline {lstlisting}{code/w.sh}{65}}
\@writefile{toc}{\contentsline {section}{\numberline {A.2}w.awk}{66}}
\@writefile{lol}{\contentsline {lstlisting}{code/w.awk}{66}}
\@writefile{toc}{\contentsline {section}{\numberline {A.3}apply.awk}{71}}
\@writefile{lol}{\contentsline {lstlisting}{code/apply.awk}{71}}
\@writefile{toc}{\contentsline {section}{\numberline {A.4}contrast.awk}{79}}
\@writefile{lol}{\contentsline {lstlisting}{code/contrast.awk}{79}}
\@writefile{toc}{\contentsline {section}{\numberline {A.5}discretize.awk}{81}}
\@writefile{lol}{\contentsline {lstlisting}{code/scripts/discretize.awk}{81}}
\@writefile{toc}{\contentsline {section}{\numberline {A.6}neighbors.awk}{84}}
\@writefile{lol}{\contentsline {lstlisting}{code/neighbors.awk}{84}}
\@writefile{toc}{\contentsline {section}{\numberline {A.7}projects.awk}{85}}
\@writefile{lol}{\contentsline {lstlisting}{code/projects.awk}{85}}
\@writefile{toc}{\contentsline {section}{\numberline {A.8}util.awk}{88}}
\@writefile{lol}{\contentsline {lstlisting}{code/util.awk}{88}}
\@writefile{toc}{\contentsline {chapter}{\numberline {B}Example Dataset and Project Descriptions}{96}}
\@writefile{lof}{\addvspace {10\p@ }}
\@writefile{lot}{\addvspace {10\p@ }}
\newlabel{sect:dataset}{{B}{96}}
\@writefile{toc}{\contentsline {section}{\numberline {B.1}NASA93 Project Descriptions}{96}}
\@writefile{toc}{\contentsline {subsection}{\numberline {B.1.1}NASA Ground Software}{96}}
\@writefile{lol}{\contentsline {lstlisting}{code/projects/ground}{96}}
\@writefile{toc}{\contentsline {subsection}{\numberline {B.1.2}NASA Flight Software}{96}}
\@writefile{lol}{\contentsline {lstlisting}{code/projects/flight}{96}}
\@writefile{toc}{\contentsline {subsection}{\numberline {B.1.3}NASA Orbital Space Plane (OSP)}{97}}
\@writefile{lol}{\contentsline {lstlisting}{code/projects/osp}{97}}
\@writefile{toc}{\contentsline {subsection}{\numberline {B.1.4}NASA Orbital Space Plane 2 (More Limited Scope)}{97}}
\@writefile{lol}{\contentsline {lstlisting}{code/projects/osp2}{97}}
\@writefile{toc}{\contentsline {section}{\numberline {B.2}NASA93 Historical Data for Defects, Effort, and Months}{98}}
\@writefile{lol}{\contentsline {lstlisting}{code/datasets/nasa93.dat}{98}}
\bibstyle{plain}
\bibdata{refs,local_refs,myrefs,thesis}
\bibcite{endres03}{1}
\bibcite{aamod94}{2}
\bibcite{andrews07}{3}
\bibcite{azzeh08}{4}
\bibcite{baker07}{5}
\bibcite{bartlett32}{6}
\bibcite{boehm81}{7}
\bibcite{boehm00a}{8}
\bibcite{Boehm96softwarecost}{9}
\bibcite{boehm00b}{10}
\bibcite{brady10b}{11}
\bibcite{brady10a}{12}
\bibcite{brooks75}{13}
\bibcite{chulani99}{14}
\bibcite{craw94}{15}
\bibcite{Dil84}{16}
\bibcite{me10f}{17}
\bibcite{fenton99}{18}
\bibcite{fenton07}{19}
\bibcite{gof95}{20}
\bibcite{me09i}{21}
\bibcite{hall03}{22}
\bibcite{harman04}{23}
\bibcite{holzmann97}{24}
\bibcite{jalali08}{25}
\bibcite{jorgensen05}{26}
\bibcite{kemerer87}{27}
\bibcite{journals/ese/KhoshgoftaarS03}{28}
\bibcite{kirkpatrick83}{29}
\bibcite{Kirsopp2002}{30}
\bibcite{me10c}{31}
\bibcite{kol93}{32}
\bibcite{kolodner83}{33}
\bibcite{leake05}{34}
\bibcite{leake96}{35}
\bibcite{Li2009}{36}
\bibcite{Lipowezky1998}{37}
\bibcite{loftus03}{38}
\bibcite{lowry10}{39}
\bibcite{Mendes2003}{40}
\bibcite{me09f}{41}
\bibcite{me07g}{42}
\bibcite{me08c}{43}
\bibcite{me07f}{44}
\bibcite{me01c}{45}
\bibcite{me09a}{46}
\bibcite{me09e}{47}
\bibcite{me05a}{48}
\bibcite{me02f}{49}
\bibcite{me10b}{50}
\bibcite{me00f}{51}
\bibcite{me10a}{52}
\bibcite{miller02}{53}
\bibcite{miyazaki94}{54}
\bibcite{mozina04}{55}
\bibcite{ruhe09}{56}
\bibcite{orrego09}{57}
\bibcite{ostrand04}{58}
\bibcite{pendharkar05}{59}
\bibcite{pfahl05}{60}
\bibcite{port08}{61}
\bibcite{norvig03}{62}
\bibcite{schank83}{63}
\bibcite{shank77}{64}
\bibcite{schulz10}{65}
\bibcite{selman96}{66}
\bibcite{shepperd07}{67}
\bibcite{shepperd97}{68}
\bibcite{She02}{69}
\bibcite{zimmermann09}{70}
\bibcite{tosun10}{71}
\bibcite{Walkerden1999}{72}
\bibcite{watson98}{73}
\bibcite{witten05}{74}
\bibcite{wolpert94}{75}
\bibcite{zhang07a}{76}