Classification of universities in Malaysia

Abstract

Universities attractiveness are measured based on its research output performance. The objective of this research is to classify 20 public universities in Malaysia in several clusters by using average linkage, K-means clustering methods and correspondence analysis. The source data was found in Buku Perangkaan 2009 from Ministry of Higher Education (MOHE) and MyMohes system. The classification of these universities are carried out using average linkage and K-means clustering and correspondence analysis is divided into 4 clusters. For clustering method, the distance measured is referred as Euclidean distance whereas for Chisquare distance, it is applied in correspondence analysis. The analyzed results showed the average linkage clustering and K-means clustering are performed at the same cluster solution that are 4 cluster solutions. Besides that, the cluster solution for correspondence analysis method is 4 and for MOHE method showed 3 cluster solutions. Clustering method, correspondence analysis method and MOHE method show different membership in each cluster.