OPTIMASI PENGKLASTERAN MENGGUNAKAN FUZZY C-MEANS PADA PESERTA IMUNISASI RUTIN DI PROVINSI JAWA TENGAH

Abstract

Health development aims to improve the ability to lead healthy lives for the community. Health programs, especially immunization, affected by the spread of Covid-19. In Central Java, the coverage of routine immunization of infants under 5 years old is low, which can increase the risk of extraordinary events. The reason of this analyzing is to give grouped results immunization coverage that is spread irregularly and knowing areas that require more attention in improving their services. The cluster formation algorithm used is Fuzzy C-Means Clustering which is a grouping technique to determine cluster members based on their membership level where the initial value is randomly selected so that local optimum occurs, then Silhouette Coefficient and Davies Bouldin Index validation are used to obtain optimal clusters. The results of grouping 35 regencies/cities in Central Java display that the optimal quantity of clusters is 3 clusters using the Euclidean distance where the highest Silhouette Coefficient is 0.5847 and the lowest Davies Bouldin Index is 0.7785. The distribution of routine immunization in Central Java Province in 2021 is quite good, but the distribution of measles vaccine between districts or cities in Central Java Province is uneven and still relatively low.