Fundamental Study of Traffic Prediction in Tokyo Bay using Machine Learning

As ships in a large bay are headed to various destination ports, there is a high possibility that these ships will collide with each other. In congested sea areas, veteran navigators predict the movement of other ships and evacuate in advance to minimize the danger of collision with other ships. Therefore, this study aimed to predict the movement of other ships, similar to veteran navigators, using big data of ship traffic flow in a bay and a machine learning system. By learning ship traffic data using the machine learning system, a model to recursively calculate the trajectories of ships was constructed.