Predicting students drop out: a case study (Technical report)

In the emerging field of educational data mining, a strong bias towards data-rich digital learning environments is the current state of affairs [2, table 2]. However, in many educational institutes a lot of regular course data will probably be more readily available. This data may also be used to support and advise students in various ways, for the better of the student as well as the institute. In this study, the situation at the Electrical Engineering department of Eindhoven University of Technology is considered. Based on experience, the department claims to be able to distinguish the potentially successful students from amongst the first year influx before the end of the first semester. To do this in an early stage is important for the student as well as for the university, but the selection is only loosely based on assumed student similarities over the years. There is no thorough analysis. Data mining techniques may corroborate and improve the accuracy of this prediction. Furthermore, data mining techniques may point out indicators of academic success that are missed until now. This study explores if and how data mining techniques can be applied in this practical situation. The techniques are applied on data that is readily available in the institution’s database. In contrast to some other studies [3], no additional data is collected to make the results of the applied techniques easily applicable. 1This is the extended technical report, accompanying the paper under the same title [1]. The report describes the data mining process in detail, including preprocessing data-related issues, and is intended to be accessible for a general public. The paper concentrates more on the results obtained and is written for domain experts. This report is the final result of an internship by the author at the Department of Computer Science, Eindhoven University of Technology, carried out as part of the Master’s program Electrical Engineering. The intership was supervised by Mykola Pechenizkiy (Assistant Professor at the Department of Computer Science) and Jan Vleeshouwers (Student Counsellor at the Department of Electrical Engineering).