Πρόβλεψη Τιμών Κρυπτονομισμάτων με Χρήση Μεθόδων Βαθιάς Μάθησης
Date Issued
February 28, 2024
Type
Μεταπτυχιακή Διπλωματική Εργασία
Abstract
This thesis explores cryptocurrency price prediction, focusing on Bitcoin due to its market dominance. Amidst the varied opinions on the future of digital currencies, this study aims to develop a reliable predictive model using deep learning techniques. Leveraging a dataset of daily Bitcoin prices from September 24, 2019, to February 6, 2024, we tested three deep learning models: ARIMA, SVM, and LSTM, to assess their forecasting accuracy. Our research finds significant performance differences among the models, with ARIMAX showing the highest precision in predicting future prices. This work contributes to the financial forecasting field by evaluating the predictive effectiveness of deep learning models on the volatile cryptocurrency market, providing valuable insights for investors, financial experts, and the scientific community.
Subjects