Predicting Bitcoin Prices Using Machine Learning

Abstract

Bitcoin, the first and most prominent cryptocurrency, has gained widespread adoption in the financial landscape. However, its extreme price volatility presents challenges for investors and businesses, making accurate price prediction a critical task. This study explores the application of machine learning techniques to forecast Bitcoin prices using historical data and macroeconomic indicators. The research evaluates four predictive models—Linear Regression, Support Vector Machines (SVM), Random Forest Regression, and Long Short-Term Memory (LSTM) networks—based on key performance metrics such as Root Mean Squared Error (RMSE), R-squared (R²), and Mean Absolute Error (MAE). The results indicate that Random Forest Regression outperforms other models in predictive accuracy, demonstrating robustness in capturing complex, non-linear relationships in Bitcoin price movements. Contrary to expectations, LSTM models, despite their ability to capture long-term dependencies, underperformed due to potential limitations in feature selection, hyperparameter tuning, and data characteristics.