Research Article
Supervised Machine Learning Algorithms for the Analysis of Ship Engine Data
@INPROCEEDINGS{10.1007/978-3-031-58053-6_7, author={Theodoros Dimitriou and Emmanouil Skondras and Christos Hitiris and Cleopatra Gkola and Ioannis S. Papapanagiotou and Dimitrios J. Vergados and Stavros I. Papapanagiotou and Stratos Koumantakis and Angelos Michalas and Dimitrios D. Vergados}, title={Supervised Machine Learning Algorithms for the Analysis of Ship Engine Data}, proceedings={Wireless Internet. 16th EAI International Conference, WiCON 2023, Athens, Greece, December 15-16, 2023, Proceedings}, proceedings_a={WICON}, year={2024}, month={5}, keywords={Supervised Machine Learning (ML) Linear Regression (LR) Ridge Regression (RR) Decision Tree (DT) Ensemble algorithms ship engine data engine decay prediction}, doi={10.1007/978-3-031-58053-6_7} }- Theodoros Dimitriou
Emmanouil Skondras
Christos Hitiris
Cleopatra Gkola
Ioannis S. Papapanagiotou
Dimitrios J. Vergados
Stavros I. Papapanagiotou
Stratos Koumantakis
Angelos Michalas
Dimitrios D. Vergados
Year: 2024
Supervised Machine Learning Algorithms for the Analysis of Ship Engine Data
WICON
Springer
DOI: 10.1007/978-3-031-58053-6_7
Abstract
Supervised Machine Learning (ML) algorithms are used for making predictions or decisions based on labeled data. In this paper, an overview about existing supervised ML algorithms is performed. In particular, the algorithms that are studied comprehend the Linear Regression (LR), the Ridge Regression (RR), the Decision Tree (DT), as well as Ensemble algorithms. Subsequently, a comparative analysis of the algorithms is performed using a dataset containing data about ship engines. Effective management of ship engines is important for their robust operation, which can then bring significant economic and environmental benefits. Inferences about the condition of engines and predictions about their performance could prove crucial for specifying optimal cruise parameters, early fault detection and timely service planning. The analysis demonstrates the strength and the weaknesses of each algorithm in terms of predicting decay factors of the ship engine by taking into consideration the data included to the aforementioned dataset.
