Data Analysis of Vessel Traffic Flow Using Clustering Algorithms

An unsupervised machine learning method-clustering, is introduced to conclude characteristics of vessel traffic flow data. A new way is found to implement data analysis in vessel traffic field using artificial intelligent technique. A similarity based algorithm, K-means, is selected in the clustering process for its simplicity and efficiency and a popular data mining tool named WEKA is chosen to execute the experiment. The result of the data mining experiment, which use the real data from an water way of Yangzi river, list the most related cluster centroids and related explanations, which show us the fact often be neglected. A conclusion that clustering is a suitable method to generalize multi-factor related regulations is made finally according to the mining result and its reasonable explanation.