Grouping of Multivariate Data by Social Welfare Indicators in North Sulawesi Province, Indonesia

Charles E. Mongi, Winsy C. D. Weku, Mans L. Mananohas

Multivariate data processing can be done by various methods in which one of the methods used to group data is cluster analysis. Grouping techniques that have been developed so far are hierarchy and non-hierarchy. With a good and appropriate clustering technique, it will give good grouping results. In this research, regencies / cities in North Sulawesi Province will be classified based on welfare indicators so that they will assist in policy making by the government in various aspects of life, especially in accelerating poverty alleviation. The objectives to be sought are the best clustering techniques based on welfare indicator data in North Sulawesi. The analytical method uses ward hierarchical techniques while the non-hierarchical method uses k-means. The grouping results using the hierarchical method resulted in 3 regencies / cities. Group one consists of 7 districts, group two consists of 2 districts and group three consists of 6 districts / cities. By using a non-hierarchical method and the number of groups are 3 produces group one consisting of 4 districts, group two consisting of 6 districts / cities and group three consisting of 5 districts.

Volume 12 | Issue 2

Pages: 1488-1492

DOI: 10.5373/JARDCS/V12I2/S20201189