Unconventional Computing for Estimating Academic Performance

One of the relevant tasks of the teaching-learning process is that of evaluation. In this sense, estimating the academic performance exhibited by the final evaluation of a student, at the end of the current semester, has become of particular interest for students, parents, educators, educative managers and researchers alike. More specifically, the current paper is focused in determining whether a first semester engineering student at the Universidad Tecnológica de Pereira, Colombia, will pass or fail the Mathematics I course. This paper proposes the use of the Gamma Classifier together with Wilson Editing to solve the problem. Results observed are competitive against classic models of pattern recognition.