Modeling Fuzzy Geographically Weighted Clustering with Flower Pollination Algorithm for Spatial Optimization and Clustering

This study aims to analyze the clustering of districts/cities in East Nusa Tenggara Province (NTT) using the Fuzzy Geographically Weighted Clustering method optimized through the Flower Pollination Algorithm (FGWC-FPA). The data consist of eight health and sanitation indicators for 2024. The analysis produced two clusters with distinct characteristics. Cluster 1 is dominated by areas with relatively higher rates of complementary feeding and good BCG immunization coverage but still shows a higher proportion of low birth weight (LBW) infants and limited access to drinking water and sanitation. Meanwhile, Cluster 2 demonstrates significant advantages in access to proper drinking water (90.37%) and proper sanitation (83.19%), as well as more optimal Hepatitis B immunization coverage. Evaluation of cluster validity using Classification Entropy (CE) and the Separation Index (SI) shows that the best configuration is obtained at m = 1.5 with c = 2, yielding the lowest CE value (0.584872) and reasonably good cluster separation (SI = 1.069092). Thus, the FGWC-FPA method is capable of producing optimal cluster partitioning and can serve as a basis for formulating more targeted health intervention strategies in NTT.