Crude Oil Price Forecasts Using Support Vector Regression and Technical Indicators

Abstract

Oil price forecasting has captured the attention of both researchers and academics because of the unique characteristics of crude oil prices and how they have a big impact on a lot of different parts of the economic value of the product. As a result, most academics use a lot of different ways to predict the future. On the other hand, researchers have a hard time because crude oil prices are very unpredictable and can be affected by many different things. This study uses support vector regression (SVR) with technical indicators as a feature to improve the prediction of the monthly West Texas Intermediate (WTI) price of crude oil. The root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE) measure how well the model is working. The RMSE was 1.5456, the MAE was 1.3219, and the MAPE was 1.9173 in the experiment. The results show that WTI crude oil prices are affected by technical indicators and get good performance that outperforms most other models that can be found.