A little more cleanup

mini_proj/report/waldo.tex

@@ -286,22 +286,19 @@
 		\end{itemize}
 
 		\emph{Accuracy} is a common performance metric used in Machine Learning,
-		however in classification problems where the training data is heavily
+		however in classification problems where the training data is heavily biased
-		biased toward one category, sometimes a model will learn to optimize its
+		toward one category, sometimes a model will learn to optimize its accuracy
-		accuracy by classifying all instances as one category. I.e. the
+		by classifying all instances as one category. I.e. the classifier will
-		classifier will classify all images that do not contain Waldo as not
+		classify all images that do not contain Waldo as not containing Waldo, but
-		containing Waldo, but will also classify all images containing Waldo as
+		will also classify all images containing Waldo as not containing Waldo. Thus
-		not containing Waldo. Thus we use, other metrics to measure performance
+		we use, other metrics to measure performance as well. \\
-		as well.
+
-		\\
+		\emph{Precision} returns the percentage of classifications of Waldo that are
-		\par
+		actually Waldo. \emph{Recall} returns the percentage of Waldos that were
-		\emph{Precision} returns the percentage of classifications of Waldo that
+		actually predicted as Waldo. In the case of a classifier that classifies all
-		are actually Waldo. \emph{Recall} returns the percentage of Waldos that
+		things as Waldo, the recall would be 0. \emph{F1-Measure} returns a
-		were actually predicted as Waldo. In the case of a classifier that
+		combination of precision and recall that heavily penalizes classifiers that
-		classifies all things as Waldo, the recall would be 0. \emph{F1-Measure}
+		perform poorly in either precision or recall.
-		returns a combination of precision and recall that heavily penalizes
-		classifiers that perform poorly in either precision or recall.
-		% Kelvin End
 
 		\section{Results} \label{sec:results}