Προβλεπτικοί αλγόριθμοι μηχανικής μάθησης σε χρηματοοικονομικά δεδομένα
Predictive machine learning algorithms in financial data
View/
Keywords
Προβλεπτικοί αλγόριθμοι ; Μηχανική μάθηση ; Χρηματοοικονομικά δεδομένα ; Machine learning ; Ενισχυτική μάθηση ; Reinforcement learning ; Limit Order Books (LOB) ; Financial dataAbstract
This work looks at how machine learning methods are used for checking and guessing stock market action,
focusing on Limit Order Book data. LOBs note as it happens the buy and sell offers for a stock giving
important info for price making which is key in high-frequency trading places.
This study used data from the Athens Stock Exchange, preprocessed and organized properly for
use by machine learning algorithms. The models used to predict future price movements were DeepLOB
and TransLOB, which have received a lot of attention in this area. Results obtained using DeepLOB for
each stock individually were very interesting. At the same time, a reinforcement learning model was built
with the objective of designing an automated trading strategy. Though the results were not exceptional, the
study underscored many challenges and great potential in making trades automatically through algorithms.
Generally, the findings imply that machine learning methods can act as great tools in financial
analysis, generating new opportunities for market prediction and making automated trading strategies.
Postgraduate Studies Programme
Κυβερνοασφάλεια και Επιστήμη ΔεδομένωνDepartment
Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα ΠληροφορικήςNumber of pages
73Language
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