<![CDATA[Cryptocurrency price prediction poses a significant challenge in the digital finance landscape due to its high volatility and complex data patterns. Traditional statistical methods often fail to capture the nonlinear and temporal dependencies inherent in cryptocurrency price movements. This study addresses this issue by evaluating and comparing the performance of three deep learning architectures, namely Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Bidirectional LSTM (Bi-LSTM). in predicting the closing prices of Bitcoin (BTC), Ripple (XRP), and Dogecoin (DOGE). The dataset was obtained from Yahoo Finance, covering the period from January 1, 2020, to April 30, 2025. The models were evaluated using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and Symmetric Mean Absolute Percentage Error (SMAPE), with a forecasting horizon of 30 days. The results of this study indicate that the LSTM model achieved the highest accuracy for Bitcoin and Ripple, with MAPE values of 2.58% and 4.33%, respectively. Meanwhile, the GRU model demonstrated the best overall performance for Dogecoin, with RMSE (0.0131), MAE (0.0084), MAPE (4.12%), and SMAPE (4.06%). On the other hand, the Bi-LSTM model exhibited the lowest performance across all tested cryptocurrencies. These findings highlight the importance of selecting an appropriate model for developing accurate cryptocurrency price prediction systems. This study contributes to the field by providing a detailed comparative analysis of model performance across cryptocurrencies with differing volatility patterns, offering insights for developing more robust and tailored predictive systems in volatile financial environments.]]>
Copyrights © 2025
