Clustering Menggunakan Self Organizing Maps (Studi Kasus: Data Perkembangan Anak di Kabupaten Bogor)

Abstract

Child development data which were gathered by the team from the Department of Family and Consumer Science, Bogor Agricultural University, require data processing to assess the characteristics of child development in Kabupaten Bogor. The purpose of this research is to implement the Self Organizing Maps (SOM) algorithm for data clustering and to obtain the characteristics from the clustering results. The data were obtained from 71 childrens at 2.5–3.4 years of age, 97 childrens at 3.5–4.4 years of age, and 126 childrens at 4.5–5.4 years of age. The data consist of four attributes: Cognitive, Language, Gross Motor, and Fine Motor. These data were the input for SOM algorithm. SOM clustering result was validated using Davies-Bouldin Index. The research shows that the clustering result for children at 2.5–3.4 years of age is 3 clusters, 3.5–4.4 years of age is 4 clusters, and 4.5–5.4 years of age is 3 clusters.