THE CLUSTER AND NETWORK ANALYSIS FOR COVID-19 CASES INCREASING IN SUB-DISTRICT OF BANDUNG

• Main Author: Gestiovani, Indah
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/55184
• Institution: Institut Teknologi Bandung

Increase cases of COVID-19 that continue to occur, especially in Bandung, make it necessary to do something to reduce that increase. By doing clustering and mapping every district in a network hopefully become a consideration in making decision of rules to reduce the increase of COVID-19 cases. This research observed the increase of COVID-19 cases in Bandung for three period of time; before, during, and after PSBB rules. This research using Euclidean and dynamic time warping distance measure. Using those distance measure, agglomerative hierarchical clustering (AHC) method is applied in each period of data. By comparing its cophenetic correlation coefficient, the best method in clustering is AHC - average linkage. In addition, this research determined which districts are the keystone of the network. It is formed through fuzzy relation by using centrality measures and principal component analysis. Using Euclidean distance, the keystone districts for for three periods in a row is Panyileukan, Bandung Kidul, and Cibiru. By using dynamic time warping distance, the keystone districts for for three periods in a row is Kecamatan Cidadap, Sukasari, and Andir.