Abstract:
The term poverty is no longer associated with the income dimension only. Poverty is a
complex problem related to other aspects of human life known as multidimensional
poverty. The high level of poverty in Indonesia (9.71% in 2021) needs to be known for
the distribution of special treatment. Cluster analysis can be used as an alternative
solution to answer these problems. This study discusses the comparison of single linkage,
complete linkage, and average linkage to multidimensional poverty indicator (education,
health, and standard of living) in districts/cities in Indonesia. This study shows that single
linkage and average linkage obtained two clusters with 513 members in cluster 1 and one
member in cluster 2, while complete linkage obtained three optimal clusters with 501
members in cluster 1, 12 members in cluster 2, and one member in cluster 3. Based on
cluster validation, average linkage is the best cluster analysis in the grouping of
districts/cities in Indonesia against multidimensional poverty indicators.