Social welfare mapping using the fuzzy C-means clustering algorithm in Bantul Regency.

Social welfare is concerned with fulfilling the social, financial, health, education, and recreational needs of individuals in society. The information about the level of welfare is very important to equalize welfare conditions between regions. This study aims to map and to determine the characteristics of the level of welfare in Bantul Regency, Special Region of Yogyakarta Province, Indonesia. The mapping is carried out based on the results of clustering using the fuzzy C-means clustering method. This study includes nine variables, those are poor households, households with female heads of households, people who are not/not yet in school, individuals with disabilities unemployed, households that do not have latrines for defecation, health workers, health facilities, and school facilities. The first six variables are grouped as negative indicators and the remaining ones are grouped as positive indicators. The validation process to get the best cluster utilizes the Partition Coefficient Index (PCI), Partition Entropy Index (PEI), and Xie Beni Index (XBI). The result shows that the best number of clusters is two with a PCI value of 0.9282, PEI value of 0.1378, and XBI value of 0.02895. The important variables that contribute to the social welfare in Bantul Regency are households headed by women, unemployed, people who are not/not yet in school, and households that do not have latrines for defecation. Most districts are included in cluster 2 with the prosperous level of welfare. It consists of eleven districts, namely Srandakan, Sanden, Kretek, Pundong, Bambanglipuro, Pandak, Pajangan, Dlingo, Pleret, Piyungan, and Sedayu. Meanwhile, cluster 1 consists of six districts with less prosperous levels of welfare, namely Bantul, Jetis, Imogiri, Bambanglipuro, Sewon, and Kasihan. Cluster 1 is characterized by all positive welfare indicators having lower percentage than cluster 2, and by negative indicators households headed by women, unemployed, people who are not/not yet in school having higher percentage than cluster 2. [ABSTRACT FROM AUTHOR]