<![CDATA[Ethereum, as one of the major crypto assets, has high price volatility, creating a need for accurate predictive models to aid investment decision-making. This study aims to implement the performance of two popular machine learning models: the Gated Recurrent Unit (GRU), a deep learning model for sequential data, and Extreme Gradient Boosting (XGBoost), an ensemble model. The data used is historical daily Ethereum price data that includes the Open, High, Low, Close, and Volume (OHLCV) features. The research method includes data pre-processing stages such as Min-Max Scaler normalization and data splitting with a ratio of 80% training data and 20% testing data. The performance evaluation of both models was measured using the Root Mean Squared Error (RMSE) and R-squared (R²) metrics. The test results show that the GRU model produces better predictions, achieving an RMSE value of 101.37 and an R² of 0.9718, while the XGBoost model obtained an RMSE value of 107.29 and an R² of 0.9656. This indicates that GRU's ability to capture temporal patterns and dependencies in time-series data is superior for Ethereum price prediction. The study concluded that the GRU model is more effective and reliable for predicting Ethereum prices than XGBoost in this study.]]>
Copyrights © 2025
