Characterization of university drop-out at UNRN using data mining. A study case

Abstract

At the National University of Río Negro (UNRN), and its Atlantic Coast Delegation in particular, it is an increasing concern for the courses corresponding to the Bachelor's Degree in Systems, the drop-out and crumbling rates observed in the first four years of the Institution. This paper describes the process of identifying the most relevant features of the problem through which, using Data Mining (DM) techniques, a college drop-out model can be obtained for the academic unit mentioned above. In order to identify the most relevant features, after processing the data we will analyze attribute projections for the expected classes or responses. The results of its application to the student data from the courses of the UNRN have been satisfactory, which allows making some recommendations aimed at reducing the percentage of students who drop put from their courses.