A Comparative Study between the Facebook Prophet Model and SARIMA for Bitcoin Price Prediction

The volatile nature of cryptocurrency markets, particularly Bitcoin, has led to an increased interest in
developing accurate forecasting models for price predictions. In this study, we conduct a comparative analysis between
two widely used time series forecasting models, the Facebook Prophet Model and SARIMA (Seasonal Autoregressive
Integrated Moving Average), to assess their effectiveness in predicting Bitcoin prices. Based on the conducted tests, the
results indicate that the Facebook Prophet model performs less favourably compared to SARIMA for daily Bitcoin
predictions. However, the Facebook Prophet model yields better results than SARIMA in monthly predictions. The
Mean Absolute Percentage Error (MAPE) for daily Bitcoin predictions is 6.7% when using the Facebook Prophet
model, whereas the MAPE for daily Bitcoin predictions with the SARIMA model is 4.5%. The MAPE for monthly
Bitcoin predictions using the Facebook Prophet model is 2.6%, while the MAPE for monthly Bitcoin predictions with
the SARIMA model is 8.4%.