REGIONAL GROUPING BASED ON POVERTY INDICATORS USING FACTOR ANALYSIS, K-MEANS CLUSTERING AND DISCRIMINANT ANALYSIS

Abstract

In 2020 Indonesia was hit by Covid-19, and one of the impacts was that the poverty rate rose. One of the goals of the SDGs is to end all forms of poverty. East Java is one of the provinces that is also affected. The purpose of this study is to describe the factors that shape Poverty Indicators in Madya districts/Cities in East Java Province, map Madya districts/Cities in East Java in 2020 based on poverty indicators, and find out the differences between the groups formed. There are 16 poverty indicator variables used in this study. The data was obtained through East Java Province in Numbers, East Java Provincial Health Statistics, East Java Provincial Education Statistics, and the website of the Central Statistics Agency. The method used is factor analysis, followed by Cluster analysis with K-Means and Discriminant Analysis. The results of the factor analysis form four factors, namely the welfare factor of education, the economic welfare factor, the factor of pln users and baduta breastfeeding, and the factor of contraceptive users (KB). Continuing with the analysis using K-Means, it produces three groups, group one is a group with moderate poverty, group two is a group with high poverty occupied by Sumenep Regency and group three is with low poverty. Followed by the Discriminant analysis, the four factors are distinguishing variables with a classification accuracy of 100 percent. The difference between this research and the research used as a reference is using non-hierarchical clusters (K-Means) while the research used as a reference uses hierarchical clusters, the results of this research analysis there are levels of poverty from the three groups formed.

Keywords : Discriminant Analysis, Factor Analysis, Poverty Indicators, K-Means Cluster, SDGs