Perbandingan Hasil Analisis Cluster dengan Menggunakan Metode Single Linkage dan Metode C-Means

  • Maria Goreti Mahasiswa Program Studi Statistika FMIPA Universitas Mulawarman
  • Yuki Novia Nasution Dosen Program Studi Statistika FMIPA Universitas Mulawarman
  • Sri Wahyuningsih Dosen Program Studi Statistika FMIPA Universitas Mulawarman

Abstract

Cluster analysis is one of the multivariate analysis which is used to classify objects into groups based on similarity of observed variables, in order to obtain the similarity of objects in the same group compared between objects of different groups. Cluster analysis is divided into two methods, they are is hierarchy method that start grouping with two or more objects that have the closest similarity and non-hierarchical method that begin with the process of determining the number of clusters in advance. This study aims is to determine whether there are differences in the results of the cluster grouping formed by using the hierarchy method, that is single linkage method, and non-hierarchical method, that is C-means method. Data, which is taken from the Environment Agency West Kutai, is data Ambient Air Quality Levels in Plantation Company in West Kutai in 2014. The results showed that based on the type of pollutants from all aleven the eleventh plantation companies have different results clusters formed from both methods which were used. With the characteristics of each cluster or groups: single linkage method for the first cluster has good air quality and its members as much as 7 companies, second Cluster both have poor air quality and its members as much as two companies and for the third Cluster have fairly  good air quality and its members as much as 2 companies. As for the method of C-means for the first cluster has good air quality and its members as many as four companies, second Cluster both have poor air quality and its members as many as four companies and third Cluster have fairly good air quality and its members as much as 3 companies. For the average value of the ratio of standard deviation in the group (Sw) and between groups (Sb) by using the method of single linkage has a smaller value that is equal to 0.022 while C-means method is equal to 0.063. Thus, in the case of the classification of the ambient air quality in plantation companies in West Kutai 2014, single linkage method better at classifying than C-means method.

Downloads

Download data is not yet available.
Published
2017-11-24
How to Cite
GORETI, Maria; NASUTION, Yuki Novia; WAHYUNINGSIH, Sri. Perbandingan Hasil Analisis Cluster dengan Menggunakan Metode Single Linkage dan Metode C-Means. EKSPONENSIAL, [S.l.], v. 7, n. 1, p. 9-16, nov. 2017. ISSN 2798-3455. Available at: <https://jurnal.fmipa.unmul.ac.id/index.php/exponensial/article/view/15>. Date accessed: 05 may 2024.
Section
Articles