Cluster visualization of student's entrance score using Smoothed Data Histograms

Student's entrance score was an indicator of student's academic ability. Bali State Polytechnics (BSP) is a vocational college in Bali. Each year BSP accepts students through three stages, namely PMDK, UMPN and TBS. Examinees who are passed the assessment then divided into several classes. The accuracy of student's distribution would affect learning outcomes. Therefore, it useful to make clustering of student's entrance score in order to used as a basis in determining the student class. The goal of this study was to show the distribution of student's academic ability based on their entrance score parameter. This study performed clustering of BSP student's entrance score then visualized the cluster become islands of students. Dataset was built from database of student's entrance scores in 2014 stage UMPN and TBS of Business Administration and International Business Management department. Student's entrance score were normalized before the clustering process. Clustering process done by Self Organizing Map method. Cluster visualization can be done using Smoothed Data Histograms method. The SDH created islands of students in accordance with the similarity of student's entrance score. The display was look more attractive with the concept of islands of students. Students with similar academic ability were shown in the same color. Certainly it will facilitate the top-level management to show the distribution of student's academic ability.