Μοντέλα πρόβλεψης για δασικές πυρκαγιές με την χρήση μεθόδων μηχανικής μάθησης
Forest fire prediction models using machine learning techniques
View/
Keywords
Μηχανική μάθηση ; Δασικές πυρκαγιές ; Machine learning ; Regression ; Classification ; Random forests ; SVMAbstract
Forest fires are a phenomenon whose frequency has been increasing rapidly in recent
years, causing particular concern to both the authorities and citizens. Greece is fighting an
uphill battle, with huge fires affecting the whole country. Machine learning has evolved
various statistical methods of data analysis to develop effective predictive models that can be
used in all kinds of applications, including forest fire prediction.
What will be presented in the context of this thesis is:
a. the general problem of monitoring data useful for fire prediction, as well as the
economic consequences of not predicting or extinguishing forest fires in time.
b. the statistical machine learning techniques that can be used to develop wildfire
prediction models using machine learning methods such as, for example, Ridge Regression,
Random Trees and Support Vector Machines.
c. the above techniques will then be applied to real fire data from the last 20 years in
Greece, in order to predict future fires based on meteorological data, ground data, etc.
Postgraduate Studies Programme
Εφαρμοσμένη ΣτατιστικήDepartment
Σχολή Χρηματοοικονομικής και Στατιστικής. Τμήμα Στατιστικής και Ασφαλιστικής ΕπιστήμηςNumber of pages
61Language
GreekCollections
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