Reducción de la dimensión de registros de evaluaciones académicas aplicando el algoritmo K-means

Abstract

In an educational environment there is a huge data quantity, this data can be analyzed, and it can be used in decision making process. Nowadays data tends to be more complex due to the size than conventional data and need dimension reduction. Educational Data M ining lets using Data Mining techniques for analyzing academic information in order to identify patterns that are not evident. This article presents results obtained in a research of a case of study where regard the academic performance of undergraduate students of Engineering of the Centro Universitario UAEM Valle de México. In the data analysis is used the Kmeans algorithm, WEKA and R Studio. We propose the use of Clustering to reduce the dimension of academic variables based on their grade registers getting during last periods then we work with some average measure of in order to predict the academic performance of a student. It is used R Studio for contrast the Clusteres obtained by WEKA.