A decomposition–integration model with dynamic fuzzy reconstruction for crude oil price prediction and the implications for sustainable development

Abstract

Grasping the future fluctuation characteristics and trend of oil prices form the basis for a deep understanding of the system mechanisms and development trends of related research fields. However, due to the complex features of the oil price, accurate prediction is very difficult to get. In order to improve the accuracy of international crude oil price predictions, a novel hybrid prediction model is proposed, that is improved on existing decomposition ensemble learning techniques by developing the Dynamic Time Warping Fuzzy Clustering method (FCM-DTW) as a new reconstruction rule. The hybrid model consists of four main steps. First, the West Texas Intermediate (WTI) crude oil spot price is decomposed into a series of relatively stable, different frequency eigenmode components (IMFs) using the adaptive noise complete integration empirical mode decomposition algorithm (CEEMDAN). FCM-DTW is then employed to reconstitute the IMFs into three sub-sequences. Subsequently, an Autoregressive Integrated Moving Average (ARIMA) model is selected according to the data characteristics of the reconstructed sequence and applied to predict the reconstructed components. Finally, a simple additive method is used to integrate the predicted results of each reconstructed component to generate the crude oil price prediction value. The results show that the prediction accuracy of the proposed hybrid model, based on dynamic time warping fuzzy clustering algorithm, is significantly better than the benchmarks considered in this paper.