Our VE implementation obtains the results given in tables
\ref{Task1Results} and \ref{Task2bResults} for the queries from tasks
1 and 2(b). (The results from task 2(a) will be discussed in section
\ref{Discuss2a}.) The \textit{Burglary}-network results precisely
match the results returned by the Enumeration solver with these
queries; the \textit{Insurance} query matches the reference
distribution given in the project handout.

\begin{table}
\begin{center}

\begin{tabular}{|l|l|}
\hline
\multicolumn{2}{|c|}{\textit{P(Burglary $|$ JohnCalls = true,
 MaryCalls = true)}} \\
\hline\hline
true  & 0.2841718353643929 \\
\hline
false & 0.7158281646356071 \\
\hline
\end{tabular}

\begin{tabular}{|l|l|}
\hline
\multicolumn{2}{|c|}{\textit{P(Earthquake $|$ JohnCalls = true,
  Burglary = true)}} \\
\hline\hline
true  & 0.0020199831098788364 \\
\hline
false & 0.9979800168901212 \\
\hline
\end{tabular}

\begin{tabular}{|l|l|}
\hline
\multicolumn{2}{|c|}{\textit{P(PropCost $|$ Age = Adolescent, Airbag =
  False, Mileage = TwentyThou)}} \\
\hline\hline
Thousand    & 0.49761258318989815 \\
\hline
TenThou     & 0.34597846229050366 \\
\hline
HundredThou & 0.13571003396327844 \\
\hline
Million     & 0.020698920556319802 \\
\hline
\end{tabular}
\end{center}

\caption{\label{Task1Results}Results of Variable Elimination Queries:
Task 1}
\end{table}

\begin{table}
\begin{center}

\begin{tabular}{|l|l|}
\hline
\multicolumn{2}{|c|}{$P(N$104 $|$ $N$41 = ``2'', $N$84 = ``1'', $N$116 =
    ``0'')} \\
\hline\hline
``0'' & 0.9429999988500375 \\
\hline
``1'' & 0.05700000114996252 \\
\hline
\end{tabular}

\begin{tabular}{|l|l|}
\hline
\multicolumn{2}{|c|}{$P(N$73 $|$ $N$152 = ``1'', $N$116 = ``0'', $N$43 =
    ``1'')} \\
\hline\hline
0 & 0.9979999999565901   \\
\hline
1 & 0.002000000043409999 \\
\hline
\end{tabular}

\end{center}
\caption{\label{Task2bResults}Results of Variable Elimination Queries:
Task 2(b)}
\end{table}