The global economy is strongly influenced by the production of crude oil, a major nonrenewable energy source. Businesses and economies all over the world are challenged by a greater degree of unpredictability due to the volatility and dynamics of crude oil prices. Several decomposition techniques, including empirical mode decomposition (EMD) and its different variants, are seamlessly incorporated. These decomposition techniques are also integrated with various machine-learning algorithms, which include the support vector machine (SVM), random forest, decision tree and artificial neural network (ANN), to build the hybrid model for crude oil prediction with intrinsic mode functions (IMF's) and residue generated from the actual West Texas Intermediate (WTI). Since the proposed hybrid model is based on the data-decomposition and supervised machine-learning algorithm, therefore the IMFs and residue component extracted from the daily closing prices of WTI are given as input features to these supervised learning techniques. Three important statistical metrics including the mean absolute error (MAE), mean absolute percentage error (MAPE), and root mean square error (RMSE) are utilized to check the prediction performance of the proposed model. The results such as the RMSE, MAE, and MAPE values of 1.446, 1.259, and 2.194 confirm that the complementary ensemble empirical mode decomposition with adaptive noise (CEEMDAN) integrated with SVM technique as a dependable and effective crude oil price forecast tool and demonstrate its improved precision. The results ensure profitability in an unpredictab |
