IMPROVEMENT OF K-MEANS ALGORITHM AND APPLICATIONS FOR STUDENTS TO CHOOSE MAJORS BASED ON CREDITS SYSTEM

Abstract

K-means algorithm is very effectively used in many applications about database
clustering. The authors applied this algorithm to professional clustering on the score data set,but the authorithms were inefficient in some cases, so the accuracy was not high. Therefore, in this paper, a clustering method on the group data set specific to each discipline was proposed. In addition, the improvement of K-Means algorithm for removing noise items was done in order to reduce the computation time of the algorithm. The result of this clustering will support students of Faculty of Information Technology, Ho Chi Minh City University of Food Industry in choosing their majors.

Keywords: K-Means, clustering, choose specialized, Euclidean distance, centroids.