Abstract:
Transshipment at sea is the exchange of cargo and supplies between two vessels that are far from their home ports. Transshipment help in offloading caught fish so that th...Show MoreMetadata
Abstract:
Transshipment at sea is the exchange of cargo and supplies between two vessels that are far from their home ports. Transshipment help in offloading caught fish so that the product can move to market quickly while fishing is still continued. However, illegal fishing has become widespread across the globe. Maritime natural resources are exploited by the widespread illegal transshipment activities which not only disrupts the natural ecological balance by selling of rare and endangered species to the sea market but slavery, trafficking, and bonded labor are some of the humongous problems that are faced today due to illegal transshipment activities as well. In this paper, we discussed the global patterns of transshipment behavior, boon and bane of using AIS data, and proposed a system integration consisting of a prediction and classification unit. The proposed prediction unit uses LSTM networks and for the classification unit, we explored both traditional machine learning methods (KNN, random forest, Logistic Regression, etc), deep learning algorithms (ANN) as well as ensemble models.
Date of Conference: 27-29 May 2022
Date Added to IEEE Xplore: 15 July 2022
ISBN Information: