Comparative Analysis of Machine Learning and LSTM Models for Solana Price Prediction Using Technical Indicators
Main Article Content
This study proposes a predictive framework for forecasting Solana cryptocurrency prices using a Long Short-Term Memory (LSTM) model and compares its performance with several machine learning approaches, including Linear Regression, Ridge Regression, Random Forest, Extra Trees, and Gradient Boosting. Historical price data were enhanced with technical indicators such as moving averages, Relative Strength Index (RSI), and Moving Average Convergence Divergence (MACD) to improve feature representation. Model performance was evaluated using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and the coefficient of determination (R²). The results show that Linear Regression achieved the best overall performance, producing the lowest error values and highest R² score, while the LSTM model exhibited the lowest predictive accuracy. These findings indicate that simpler machine learning models can outperform deep learning approaches when the dataset is limited and the features are highly informative. Furthermore, feature importance analysis reveals that fundamental price variables such as Close, Open, High, and Low contribute more significantly to prediction performance than complex technical indicators. Overall, the study highlights the importance of aligning model complexity with data characteristics and demonstrates that effective feature engineering plays a critical role in improving cryptocurrency price prediction accuracy.