Μέθοδοι μηχανικής μάθησης στη ναυτιλία
Machine learning applications on maximum wave height for shipping and maritime
View/
Keywords
Μηχανική μάθηση ; Γραμμική παλινδρόμηση ; Λογιστική παλινδρόμηση ; Νευρωνικά δίκτυα ; Εφαρμογές προσομοίωσης θαλάσσιου κυματισμούAbstract
This dissertation aims to give an introduction to Machine Learning’s logic notions and present its application in real-life Maritime case studies. It begins with the definition and taxonomy of Machine Learning, accompanied by its advancements through time. It then provides the computational background and fundamentals of four Supervised Machine Learning algorithms: Linear Regression, Polynomial Regression, Logistic Regression, and Neural Network. Both advantages and disadvantages are presented for all methods. The selection of MATLAB’s usage is explained, before proceeding to three applications based on real-life maritime occurrences, appearing in the Aegean and Ionian Seas. A Linear Regression model estimates the maximum wave height, a Logistic Regression model decides if a prohibition of sailing should be issued by port authorities, and a Neural Network characterizes, through wave heights, the sea condition based on the Douglas Sea Scale. The dissertation ends with the conclusions that are derived from said applications.
Postgraduate Studies Programme
ΔΠΜΣ στη Διοίκηση στη Ναυτική Επιστήμη και ΤεχνολογίαDepartment
Σχολή Ναυτιλίας και Βιομηχανίας. Τμήμα Ναυτιλιακών ΣπουδώνCorporate Department
Σχολή Ναυτικών ΔοκίμωνNumber of pages
111Language
EnglishCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα