ANALISIS MULTIDIMENSIONAL SCALING DAN CLUSTER FUZZY C-MEANS PADA DATA INDIKATOR KESEJAHTERAAN RAKYAT DI INDONESIA

Abstract

The Covid-19 pandemic has an impact on the level of people's welfare in regional segmentation. Graphical and clustering can be done to determine the level of welfare of the people in Indonesia during the pandemic so that segmental policies can be taken. This study uses the Central Bureau of Statistics data, namely indicators of the welfare of the people in Indonesia which consists of GRDP value, per capita expenditure, percentage of poor people, level of open space, and average length of schooling. Graphical analysis using multidimensional scaling produced 2 groups of provinces consisting of 30 and 4 provinces and of regencies/cities consisting of 494 and 20 regencies/cities. Validation of multidimensional scaling of STRESS values close to 0 and R square equal to 1. Clustering using fuzzy C-Means for welfare status by province has 2 clusters consisting of 30 and 4 provinces and by regencies/cities consisting of 495 and 19 regencies/cities. Comparison of each variable obtained cluster 2 is more prosperous than cluster 1. Cluster validation using MPC and FSI is close to 1.