Anàlisi de components principals i discriminant lineal: una aplicació a resultats acadèmics

Abstract

[en] Aiming to give an introduction of Data analysis and support the innovation in teaching project of the University of Barcelona which has the purpose of creating an intelligent support system for both the head of studies and the tutors, this work studies the Principal Components Analysis and the Linear Discriminant Analysis; from a theoretical basis in order to understand its mechanics and from a practical implementation, applying them to academic results in order to extract information regarding the abandonment of university studies. The sample used are the degree in Mathematics, Computer Science and Law, for every one of which we considered the grades of the first course taking into account the students that abandoned separated and together with the students that finished. The project has mainly been implemented in Python. For visualization purposes we also used R.