<![CDATA[The Sunda Strait is a strategic shipping lane with a high density of vessel traffic, which increases the potential risk of maritime accidents. Therefore, analyzing traffic density patterns is essential to support safer and more efficient navigation management. This study aims to identify vessel traffic density patterns using machine learning-based clustering techniques, as well as to compare the performance of the K-Means and DBSCAN algorithms. The data used are secondary data obtained from VTS Merak, consisting of daily vessel counts categorized based on Traffic Separation Scheme (TSS) routes, namely Passing North, Passing South, Crossing West, Crossing East, and Vessel Not Using TSS. The data were processed through preprocessing stages, including data cleaning and normalization using the Min-Max Scaling method. The analysis was conducted by applying K-Means and DBSCAN algorithms and evaluated using the Silhouette Score and Davies-Bouldin Index. The results indicate that DBSCAN is more effective in identifying complex density patterns, forming density-based clusters, and detecting anomalies in vessel trajectories. In contrast, K-Means produces more structured clusters but is less flexible in handling irregular data patterns. Therefore, DBSCAN is considered superior for analyzing vessel traffic density in the Sunda Strait and has strong potential to support decision-making in improving maritime safety.]]>
Copyrights © 2026
