Update on Overleaf.

main.tex

@@ -73,15 +73,16 @@ \section{Analysis}
 \subsection{Data Exploration}
 The dataset used is composed by level I order book data from PETR4, a stock traded at BMFBovespa Stock Exchange. Includes 45 trading sessions from 07/25/2016 to 09/26/2016. I will use the first day to create the scalers of the features used, that I shall explain. Then, I will use four other days to train and test the model, and after each training session, I will validate the policy found in an unseen dataset from the subsequent day.
 
-In the figure below can be observed how the market behaved on the days that the out-of-sample will be performed. In this charts are plotted the number of cents that an investment of the same amount of money in PETR4 and BOVA11 would have varied on these market sessions. BOVA11 is an ETF\footnote{Source: \url{https://en.wikipedia.org/wiki/Exchange-traded_fund}} that can be used as a proxy to the [Bovespa Index](https://goo.gl/TsuUVH), the Brazilian Stock Exchange Index. As can seem, PETR4 was relatively more volatile than the rest of the market.
+In the figure \ref{fig:market} can be observed how the market behaved on the days that the out-of-sample will be performed. In this charts are plotted the number of cents that an investment of the same amount of money in PETR4 and BOVA11 would have varied on these market sessions. BOVA11 is an ETF\footnote{Source: \url{https://en.wikipedia.org/wiki/Exchange-traded_fund}} that can be used as a proxy to the Bovespa Index
+\footnote{Source: \url{https://goo.gl/TsuUVH}}, the Brazilian Stock Exchange Index. As can seem, PETR4 was relatively more volatile than the rest of the market.
 
 \begin{figure}[ht]
 \centering
-\includegraphics[width=0.65\textwidth]{figures/market.png}
+\includegraphics[width=0.75\textwidth]{figures/market.png}
 \caption{\label{fig:market}Market and Stock Variation.}
 \end{figure}
 
-The data used in simulations was collected from Bloomberg. There are 45 files, each one has 110,000 rows on average, resulting in 5,631,273 rows at total and almost  230 MB of information. In the figure \ref{tab:original_data} is the structure of one of them:
+The data used in the simulations was collected from Bloomberg. There are 45 files, each one has 110,000 rows on average, resulting in 5,631,273 rows at total and almost  230 MB of information. In the figure \ref{tab:original_data} is the structure of one of them:
 
 \begin{table}[ht!]
 \centering