The Clustering of Households in Madura Based on Factors Affecting Their Ingestion of Clean Water Using Similarity Weight and Filter Method

Abstract

Clean Water and Sanitation is one of SDGs’ indicators that relates to human’ demand for clean water. Three of four regencies in Madura Island reportedly have suffered in drought, thus it leads this research to fulfill Madura people need of water. Madura Island has 3097 households in need of water. However, not all households could fetch their need. This research aims to classify the households of Madura Island regarding factors which affect their ingestion of clean water using cluster analysis. There are clustering numerical data and categorical data. Therefore, this research uses Similarity Weight and Filter Method. SWFM is one of clustering mix methods in which there are clustering numerical, using hierarchical ward, and clustering categorical, using k-modes. To analyze the clustering numerical data, there are 3 variables and it gains two optimum groups by using ward method with pseudo-F 1001,172. Clustering categorical analysis uses 6 variables with k-modes and gains three groups and SWFM gains five groups. Five groups are selected because they produced the smallest ratio 0,006627 in the group.