<![CDATA[As an energy source, crude oil price forecasting is crucial for a country's policy decisions. However, crude oil price volatility is influenced by many factors, one of which is geopolitical conditions. During the Iran-United States conflict, which led Iran to close the Strait of Hormuz, an international oil distribution route, oil prices reached a spike that could not be considered by univariate time series models. This study proposes a framework that combines the Autoregressive Integrated Moving Average (ARIMA) model with a machine learning algorithm. ARIMA is used to predict the dollar index and divided into two models: normal conditions and wartime conditions. Furthermore, the Decision Tree, Random Forests, and Extreme Gradient Boosting (XGBoost) machine learning models are used to predict oil prices based on historical information on oil prices and the dollar index, as well as the closure and opening status of distribution routes. The results show that this integration is capable of predicting oil prices with MAE less than 10 in this geopolitics situation.]]>
Copyrights © 2026
