To build a reliable classification model, the
CRISP-DM methodology (Cross-Industry
Standard Process for Data Mining) [8] is
adopted. The methodology consists mainly of
five steps: Collecting the relevant features of the
problem under study, preparing the data,