Electronic Theses and Dissertation

Indeks Pembangunan Manusia (IPM) mencerminkan perkembangan sosial dan ekonomi daerah, dimana kesehatan memainkan peran penting termasuk pengendalian penyakit menular. Indonesia, sebagai wilayah tropis memiliki potensi besar untuk munculnya penyakit menular baru serta munculnya kembali penyakit menular yang awalnya masih mampu untuk dikendalikan, yang menjadi tantangan besar bagi sistem kesehatan. Penelitian ini bertujuan untuk memperoleh jumlah cluster optimal untuk kabupaten/kota di Indonesia berdasarkan penyakit menular menggunakan Bayesian Information Criterion dan memvalidasi kualitas cluster dengan Davies Bouldin Index (DBI), Dunn Index (DI), dan Silhouette Index (SI), serta untuk mengelompokkan semua kabupaten/kota di Indonesia menggunakan metode Gaussian Mixture Model dan memetakan distribusi penyakit menular berdasarkan hasil pengelompokan tersebut. Penelitian ini menggunakan data penyakit menular di 514 kabupaten/kota di Indonesia pada tahun 2019-2022 dengan 4 variabel yaitu pneumonia, diare, malaria, dan DBD, yang diperoleh dari BPS. Hasil penelitian menunjukkan 3 jumlah cluster optimal berdasarkan nilai BIC terbaik, yaitu model 1 (5 cluster, EEV), model 2 (6 cluster, EEE), dan model 3 (5 cluster, EEE). Model 3 dipilih sebagai model terbaik berdasarkan nilai BIC sebesar -7657.072, DBI terkecil, DI sebesar 1.168758, dan SI tertinggi. Cluster 1 didominasi kasus diare di Sumatera Utara (kuning). Cluster 2 didominasi kasus diare dan malaria di Papua (biru). Cluster 3 didominasi kasus diare, pneumonia, dan DBD di Sulawesi Selatan dan Aceh (merah muda). Cluster 4 didominasi kasus diare, pneumonia, dan DBD di Jawa Timur, Jawa Tengah dan Jawa Barat (hijau). Cluster 5 didominasi kasus malaria dan DBD di Papua dan Papua Barat (jingga).

The Human Development Index (HDI) reflects the social and economic development of a region, where health plays a significant role, including the control of infectious diseases. Indonesia, as a tropical region, has a high potential for the emergence of new infectious diseases and the re-emergence of previously controlled diseases, posing significant challenges to the healthcare system. This study aims to determine the optimal number of clusters for regencies/cities in Indonesia based on infectious diseases using Bayesian Information Criterion (BIC) and to validate the quality of cluster using the Davies-Bouldin Index (DBI), Dunn Index (DI), and Silhouette Index (SI). Additionally, this study seeks to cluster regencies/cities in Indonesia using the Gaussian Mixture Model and map the distribution of infectious diseases based on the clustering results. The study uses infectious diseases data of 514 regencies/cities in Indonesia form the year 2019-2022, with 4 variables which are pneumonia, diarrhea, malaria, and dengue fever (DF), obtained from BPS. The results indicate 3 optimal cluster models based on the best BIC values which are model 1 (5 cluster, EEV), model 2 (6 cluster, EEE), and model 3 (5 cluster, EEE). Model 3 was selected as the best model, based on the BIC value of -7657.072, the smallest DBI, a DI of 1.168758, and the highest SI. Cluster 1 is dominated by diarrhea cases in North Sumatra (yellow). Cluster 2 is dominated by diarrhea and malaria cases in Papua (blue). Cluster 3 is dominated by diarrhea, pneumonia, and DF cases in South Sulawesi and Aceh (pink). Cluster 4 is dominated by diarrhea, pneumonia, and DF cases in East Java, Central Java, and West Java (green). Cluster 5 is dominated by malaria and DF in Papua and West Papua (orange).