Πρόβλεψη κατανάλωσης καυσίμου πλοίων με τεχνικές μηχανικής μάθησης

Αλεξιάς, Παύλος; Alexias, Pavlos

dc.contributor.author	Αλεξιάς, Παύλος	el
dc.contributor.author	Alexias, Pavlos	en
dc.date.accessioned	2022-11-02T07:04:07Z
dc.date.available	2022-11-02T07:04:07Z
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/56056
dc.identifier.uri	http://dx.doi.org/10.26240/heal.ntua.23754
dc.rights	Default License
dc.subject	Τεχνητή Νοημοσύνη	el
dc.subject	Ναυτιλία	el
dc.subject	Γραμμική Παλινδρόμηση	el
dc.subject	Νευρωνικά Δίκτυα	el
dc.subject	Πολυωνυμική Παλινδρόμηση	el
dc.subject	Artificial Intelligence	en
dc.subject	Maritime	en
dc.subject	Linear Regression	en
dc.subject	Polynomial Regression	en
dc.subject	Neural Network	en
dc.title	Πρόβλεψη κατανάλωσης καυσίμου πλοίων με τεχνικές μηχανικής μάθησης	el
heal.type	bachelorThesis
heal.classification	Μηχανική Μάθηση	el
heal.language	el
heal.access	campus
heal.recordProvider	ntua	el
heal.publicationDate	2022-07-14
heal.abstract	The goal of this thesis is to calculate the fuel needs of a vessel by using machine learning algorithms. More specifically, by using linear regression as well as polynomial regression, in order to accurately predict the dependent variable of fuel consumption was the main goal of this thesis. As independent variables other metrics were considered, such as the deadweight of the vessel and the power of the vessel's machine. Lastly, in order to avoid the linearity that the above algorithms assume, a neural network was created in order to cross check the results. The inputs of the neural network are the same as the independent variables of the machine learning algorithms. Finally, a Graphical User Interface was created, in order for the user to easily and quickly interact with the models. The user, through the use of simple components, can insert the specifications of the vessel that he is interested in. As an output, the user sees the prediction of fuel consumption - both from the linear and the neural network solutions - as well as graphs containing information with vessels similar to the one the user inserted.	en
heal.advisorName	Ασκούνης, Δημήτριος	el
heal.committeeMemberName	Ασκούνης, Δημήτριος	el
heal.committeeMemberName	Ψαράς, Ιωάννης	el
heal.committeeMemberName	Δούκας, Χρυσόστομος	el
heal.academicPublisher	Εθνικό Μετσόβιο Πολυτεχνείο. Σχολή Ηλεκτρολόγων Μηχανικών και Μηχανικών Υπολογιστών. Τομέας Ηλεκτρικών Βιομηχανικών Διατάξεων και Συστημάτων Αποφάσεων	el
heal.academicPublisherID	ntua
heal.numberOfPages	58 σ.	el
heal.fullTextAvailability	false