<![CDATA[The high volatility of cryptocurrency assets such as Bitcoin and Ethereum poses significant challenges for producing accurate and transparent price predictions. This instability is driven by factors including trading volume, market sentiment, and global economic conditions, requiring predictive models that not only forecast price movements but also explain the contribution of each variable. This study analyzes the factors influencing Bitcoin and Ethereum price prediction using a Long Short-Term Memory (LSTM) model integrated with Explainable Artificial Intelligence (XAI) through SHAP (SHapley Additive exPlanations). The dataset was obtained from Yahoo Finance covering January 2022–December 2024. Model evaluation using MAE, MSE, RMSE, and the coefficient of determination (R²) indicates that the model captures the main price movement trends, although it shows limitations in representing extreme short-term fluctuations due to the high volatility of cryptocurrency markets. Autocorrelation Function (ACF) analysis of residuals suggests that the primary temporal patterns have been effectively learned by the model. SHAP analysis identifies Low, High, and short-term Moving Average (MA7) as the most influential features. Overall, the integration of LSTM and XAI enables interpretable cryptocurrency price prediction, supporting a more transparent and data-driven understanding of market dynamics and investment decision-making.]]>
Copyrights © 2026
