Use of Unsupervised Clustering to Characterize Graduate Students Profiles based on Educational Outcomes

The identification of profiles and typologies of students plays interesting role in adapting educational approaches and improving academic outcomes. It is with this perspective that we will show, in this work, how unsupervised learning techniques can be applied to educational data for the extraction of typologies and student profiles. We will also implemented this Clustering analysis using K-means algorithm and R programming language, in order to identify homogeneous groups of students, according to their academic performance in combination with the length of studies of the bachelor program. The dataset used in this study consists of student's official grades for the six semesters and the final grade of the bachelor degree. The approach presented in this study will enrich the understanding of different learning characteristics of graduatestudents and could be used to adapt teaching approaches and strategies according to the identified student profiles.