PERBANDINGAN METODE DALAM ANALISIS CLUSTER UNTUK MENGELOMPOKKAN KABUPATEN/KOTA DI JAWA TENGAH
Abstract
Welfare is a dynamic condition. Efforts that can be done in framework of implementation development programs to be right on target in making decisions and development strategies in Central Java that is identification welfare characteristics of each region. One way of identification is
use cluster analysis to group objects with similar characteristics. The main purpose of this final project is to know the result of 3 clusters formed between hierarchy method that is Single Linkage, Complete Linkage, Average Linkage and non-hierarchy method, that is K-Means. Next to find out the best method by comparing standard deviation value in the cluster () and between clusters ().
That has smallest ratio () to () is the best method. Results showed that Single Linkage was the best method and formed 3 clusters, the first cluster was 33 regency / cities, the second cluster was 1 regency / city, and the third cluster 1 regency / city with ratio standard deviation 0.224599.
Interpretation the first cluster have advantages low population density, low average per capita expenditure, and ownership status of his own residence. The second cluster has advantages of low open unemployment, low poverty rate, low malnutrition, and average school duration is high.
While the third cluster has advantages in all sector.
Keywords: Welfare, Cluster Analysis, Hierarki, Non-Hierarki
use cluster analysis to group objects with similar characteristics. The main purpose of this final project is to know the result of 3 clusters formed between hierarchy method that is Single Linkage, Complete Linkage, Average Linkage and non-hierarchy method, that is K-Means. Next to find out the best method by comparing standard deviation value in the cluster () and between clusters ().
That has smallest ratio () to () is the best method. Results showed that Single Linkage was the best method and formed 3 clusters, the first cluster was 33 regency / cities, the second cluster was 1 regency / city, and the third cluster 1 regency / city with ratio standard deviation 0.224599.
Interpretation the first cluster have advantages low population density, low average per capita expenditure, and ownership status of his own residence. The second cluster has advantages of low open unemployment, low poverty rate, low malnutrition, and average school duration is high.
While the third cluster has advantages in all sector.
Keywords: Welfare, Cluster Analysis, Hierarki, Non-Hierarki
Full Text:
PDFRefbacks
- There are currently no refbacks.