How Predicting The Academic Success of Students of the ESPAM MFL?: A Preliminary Decision Trees Based Study

The success of higher education institutions can be measured by the students performance. Identifying preferences, factors or behaviours that increase the academic success rate of students is helpful since it can aid educational decision makers to adequately plan actions to promote their success outcomes. In this paper, we determine academic success of students of the ESPAM MFL through decision trees based algorithms as a preliminary approach. We use three built classifiers: C5.0, Random Forest and CART which are applied on a dataset with 1086 instances corresponding to personal and academic information about professionalizing subjects of students from the Computer Science Career. We train and test the algorithms considering the academic success as a multi-class classification problem, where each student has a performance mutually exclusive: Acceptable, Good, Excellent. We evaluate the algorithms verifying their classification capacity through performance metrics for classification problems. Finally, the CART algorithm was considered as the best algorithm based on its performance. The highest classification metrics values achieved by it are accuracy = 52%, precision=49% and recall=53%.