deletions | additions       diff --git a/experiments.tex b/experiments.tex  index 872a595..6c13fcf 100644  --- a/experiments.tex  +++ b/experiments.tex 
...
\section{Experimental evaluation}  \label{sec:exp}  In this section we describe the experiments we have done, starting with the data gathering process, the training of the models and the results of their evaluation.  \subsection{Data set and processing}  The data set that was obtained from 20 receipts that were manually annotated with the position of each character in them. In total there are 7045 characters. There are 74 different characters, including digits, uppercase and lowercase letters and punctuation.  
...
\end{tabular}  \end{table}  Table \ref{table:seg_values} contains the average, maximum and minimum values obtained for the F1 measure measure\cite{fawcett2006introduction}  of the character segmentation problem. The F1 measure is used instead of the accuracy because the two classes are imbalanced: there are 11475 data points which indicate a segmentation point, while there are 19402 points which are not segmentation points, almost twice as many. \begin{table}[h]  \caption{The F1 score for the character segmentation experiment}