| Autor: |
Magnussen, Búgvi Benjamin, Bläser, Nikolaj, Jensen, Rune Møller, Ylänen, Kenneth |
| Jazyk: |
angličtina |
| Rok vydání: |
2021 |
| Zdroj: |
Magnussen, B B, Bläser, N, Jensen, R M & Ylänen, K 2021, Destination Prediction of Oil Tankers Using Graph Abstractions and Recurrent Neural Networks . in Computational Logistics : 12th International Conference, ICCL 2021 Enschede, The Netherlands, September 27–29, 2021 Proceedings . vol. 13004, Springer, Lecture Notes in Computer Science, pp. 51-65 . |
| Popis: |
Predicting the destination of vessels in the maritime industryis a problem that has seen sustained research over the last few yearsfuelled by an increase in the availability of Automatic Identification System(AIS) data. The problem is inherently difficult due to the nature ofthe maritime domain. In this paper, we focus on a subset of the maritimeindustry - the oil transportation business - which complicates the problemof destination prediction further, as the oil transportation marketis highly dynamic. We propose a novel model, inspired by research ondestination prediction and anomaly detection, for predicting the destinationport- and region of oil tankers. In particular, our approach utilises agraph abstraction for aggregation of global oil tanker traffic and featureengineering, and Recurrent Neural Network models for the final port- orregion destination prediction. Our experiments show promising resultswith the final model obtaining an accuracy score of 41% and 87.1% on adestination port- and region basis respectively. While some related worksobtain higher accuracy results - notably 97% port destination predictionaccuracy - the results are not directly comparable, as no related literaturefound deals with the problem of predicting oil tanker destination ona global scale specifically. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
