Clustering human development index data with gravitational search algorithm-fuzzy 4-means (GSA-F4M)
Dewi Syifaur Rohmah, Susiana Sari, K. Vika Yugi · AIP conference proceedings · 2021
Human Development Index (HDI) is the indicator which measures the efforts of the human life’s successful quality in Indonesia. The description of Human development’s condition is used to evaluate the achievements of human development in each region in Indonesia, thus the analysis is needed. The analysis of the description can be solved using the clustering method which to group each provinces in Indonesia based on four HDI indicators. Fuzzy C-Means is one of clustering algorithm which is commonly used and has a high accuracy level. FCM has a limitation that is easily trapped in minimum local conditions when calculating the objective function in determining the the center of clusters, and causing the result is not the lowest value of the solution sets. The Gravitational Search Algorithm (GSA) algorithm approach is used so that the results obtained are optimum globally. The use of the GSA algorithm in FCM aims to optimize objective functions so as to overcome minimum local problems and obtain optimal cluster results. This research aims to apply GSA-FCM to cluster HDI data in Indonesia. The results of this research are the four clusters, the first cluster consists of 5 provinces, the second cluster has 9 provinces, the third cluster has 14 provinces, and the fourth cluster has 6 provinces.