DATA MINING IN GRADUATES BY USING K-MODES CLUSTERING AND CORRESPONDENCE ANALYSIS (A CASE STUDY IN PRESIDENT UNIVERSITY)

Abstract

In order to acknowledge the real tuition condition and own the comparative strength, President University should conduct a quantitative research. With the large amount of data, the powerful research methods K-modes clustering and correspondence analysis cope with the real academic data. K-modes cluster the data into several clusters according to the similarity of the attributes. The k-modes algorithm uses a simple matching dissimilarity measure to deal with categorical objects, replaces the means of clusters with modes, and uses a frequency-based method to update modes in the clustering process to minimize the clustering cost function. Moreover, the clustering result is analyzed by applying the capability of correspondence analysis to deal with frequency data, and correspondence analysis works with data that may not meet the restrictions on data necessary for other statistical analyses. The interpretive strength of correspondence analysis is its representation of low-dimensional solutions in graphical displays, which permit the researcher to make comparisons between variables, and variables in their relative placement in common low dimensional space. The correspondence analysis demonstrates the clustering result in graphical display. Consequently, the real condition of the students exhibits intuitively, as well as the recommendation is put forward to improve the condition.