Competency test clustering through the application of Principal Component Analysis (PCA) and the K-Means algorithm

Abstract

Abstract The implementation of the Competency Test at the LSP institution in higher education is an effort to ensure that students have abilities in certain fields according to predetermined competency standards. Education providers are required to always strive to improve the quality and quality of education with the aim that the student’s academic performance will always improve. From the results of observations made in the research location, it was found a problem with the high number of failures in the implementation of the competency test. This study aims to conduct cluster analysis of the data from the implementation of competency tests using Machine Learning techniques through the application of Principal Component Analysis (PCA) and K-Means Algorithm, through several stages in the form of data collection, data cleaning, data transformation, data modeling and experimentation. This study resulted in grouping the results of competency tests which were divided into 3 clusters, namely cluster 1 as much as 38%, cluster 2 as much as 32% and cluster 3 as much as 30%.