25_additional.tex

\section{Additional Features}
In order the make the predictions more precise we decided to improve the feature matrix with additional features.
We took into consideration which data could be meaningful for the frequency of bike rentals. 
We came to the conclusion that the following 4 features should be added:
\begin{itemize}
\item Age
\item Events
\item Earnings
\item Poitical Attitude
\end{itemize}
We considered the first point because we assumed that there is a specific age group which uses rental bikes more than others. We conclude from that, if we find stations where a a lot of people within this specific age group lives, we will have a higher frequency. Therefore we investigated in the target group of Santander Rental Bikes and found out that their main target group depicts of people between the age of 16 and 54 \cite{Santander}. Moreover approximately two third of their users are male. That is why we searched for data which gives us the total number of ages as well as the gender for each districct in London where Santander provides rental station data. \\\\
Appropriate data was found at the \emph{London Datastore} where data about London's population is freely provided as an excel document. To use this data we first extracted only the district data we are interested in.  
\subsection{Data Research}

\subsection{Data Preparation}

\subsection{Data Evaluation}