Analysis of the factors affecting the severity of marine accidents occurring at anchorage areas with machine learning algorithms

Anchorage areas are strategically important sea zones where ships conduct maintenance and repair activities, carry out loading and unloading operations, and temporarily wait before or after voyages. However, various studies in the scientific literature and reports by international maritime organizations highlight that these areas are also high-risk zones where a significant number of maritime accidents occur. The main objective of this study is to analyze the factors influencing the severity of maritime accidents that take place in anchorage areas and to contribute to a better understanding of safety risks in these zones. To this end, 186 accident reports from different ports and coastal regions around the world were thoroughly examined, and a comprehensive dataset consisting of both qualitative and quantitative data was constructed. This dataset was analyzed using advanced statistical and probabilistic methods, including Logistic Regression and Bayesian Network Modeling. The results of the study reveal the key factors that increase the severity of accidents in anchorage areas. The findings provide valuable insights for ship crews, port authorities, maritime administrations, and policymakers, contributing to the development of more effective and targeted preventive strategies to reduce accident risks in anchorage zones.