A novel hybrid model based on deep learning and error correction for crude oil futures prices forecast

Junhao Wu, Jinghan Dong, Zhaocai Wang, Yuan Hu and Wanting Dou

Resources Policy, 2023, vol. 83, issue C

Abstract: Energy is a crucial basis for ensuring people's life quality and advancing social and economic development. Whereas, the crude oil futures prices are influenced by several factors such as energy policies, which shows strong non-linearity and high volatility. Accurately predicting price changes of crude oil futures is a challenging task. In this study, a new hybrid model, namely VILIG based on the "decomposition-prediction-ensemble-error correction" framework, is constructed to predict the closing prices of West Texas Intermediate (WTI) and Brent crude oil futures prices. First, variational modal decomposition (VMD) is applied to decompose the original price dataset into several sub-models. Then, each sub-model is used as an input to the Long-short term memory network (LSTM) and an improved golden jackal optimization (IGJO) algorithm is proposed to optimize the hyperparameters of the LSTM. Subsequently, an improved fully ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) is introduced to perform a secondary decomposition of the error sequence and gate recurrent unit (GRU) is used to predict the decomposed error sequences separately. Finally, the prediction outcomes are obtained by linear aggregation. In addition, based on the point prediction results, this study presents kernel density estimation (KDE) for interval estimation to perform uncertainty prediction. The empirical results show that the VILIG model outperforms all the currently available major models in terms of quantitative evaluation metrics. The model can provide a data base for decision making for crude oil futures investors and market regulators.

Keywords: Crude oil futures prices; Gated recurrent unit; Hybrid model; Improved golden jackal algorithm; Long-short term memory network; Variational modal decomposition (search for similar items in EconPapers)
Date: 2023
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (2)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0301420723003136
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:jrpoli:v:83:y:2023:i:c:s0301420723003136

DOI: 10.1016/j.resourpol.2023.103602

Access Statistics for this article

Resources Policy is currently edited by R. G. Eggert

More articles in Resources Policy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().