Classification of Academic Achievement in Upper-Middle Education in Veracruz, Mexico: A Computational Intelligence Approach

Due to current technological development, there are accumulations of data from various sources, susceptible of being analyzed and interpreted. Particularly, in the education field, the analysis of data generated in educational environments, such as the academic achievement of students who reach the end of their studies at different educational levels, becomes relevant. This research presents the academic achievement of upper-middle education students in the state of Veracruz, Mexico, based on the type of school they belong to (State, Federal or Private). For this, the PLANEA-2017 database (National Plan for the Evaluation of Learning) was used, as well as computational intelligence algorithms, such as random forests and support vector machines. As a main result, it was obtained that the model produced by the random forests presented a better classification accuracy, and the most relevant variables to classify academic achievement by type of schools evaluated were the total number of students enrolled; the level of academic performance achieved by students in Language and Communication, highlighting the excellent (IV) and insufficient (I) level, respectively; and the number of students tested in Mathematics.