Towards Freshmen Performance Prediction

In this paper, we deal with freshmen performance prediction. We analyze data from courses offered to students at Faculty of Informatics, Masaryk University. We supposed that the success rate of our predictions increases when we omit freshmen from our experiments as we have no study-related data about them. However, we disproved this hypothesis because there was generally no significant difference in prediction of freshmen and non-freshmen students. We also presented the attributes that were important for freshmen performance prediction.