Explainable and Robust Oil Price Forecasting Using Hybrid Deep Learning and SHAP Interpretability: A Long-Term Analysis of WTI Prices from 1986 to 2025

Onur Polat; Neslihan Parlak; Esra Söğüt; Eda Fendoğlu; Muammer Türkoğlu

Explainable and Robust Oil Price Forecasting Using Hybrid Deep Learning and SHAP Interpretability: A Long-Term Analysis of WTI Prices from 1986 to 2025

Abstract

Since the Industrial Revolution, crude oil has become one of the most important fundamental determinants shaping the global economy. Crude oil is an indispensable input for strategic sectors such as petrochemicals, industrial production, transportation, defense, and manufacturing. Fluctuations and sudden changes in crude oil prices profoundly affect not only energy markets but also global financial stability and political decision-making processes. Therefore, accurately predicting crude oil prices is of strategic and great importance for the stability of energy markets, the sustainability of energy policies, and economic planning. In this study, a hybrid deep learning model combining convolutional neural networks (CNN), bidirectional long-short term memory (BiLSTM), and multi-head attention mechanisms was developed and analyzed using data from January 1986 to March 2025 to forecast crude oil prices. The proposed model uses multi-scale convolutional layers to capture patterns at different scales, a BiLSTM structure to learn time series dependencies, and an attention mechanism to highlight critical information. The dataset consists of West Texas Intermediate (WTI) crude oil spot prices along with macroeconomic and energy market indicators such as exchange rates, industrial production index, consumer price index, producer price indices, and global real economic activity. The model was tested using 5-fold cross-validation and obtained values of 0.0720 MAE, 0.0966 RMSE, and 0.9905 R². Furthermore, SHAP (SHapley Additive Explanations) analysis was used to determine that variables such as jet fuel, gasoline, propane, and diesel price indices contributed most significantly to the model predictions. The results show that the proposed model successfully captures the nonlinear dynamics of oil prices and offers higher accuracy compared to existing methods. The findings obtained in this study could be critical for investors, policymakers, and governments in making strategic decisions and developing profitable policies related to global crude oil prices through the implementation of the proposed model.

Keywords

References

[1] H. Guliyev and E. Mustafayev, "Predicting the changes in the WTI crude oil price dynamics using machine learning models," Resources Policy, vol. 77, 102664, March 2022. doi: 10.1016/j.resourpol.2022.102664
[2] X. Li, W. Shang and S. Wang, "Text-based crude oil price forecasting: A deep learning approach," International Journal of Forecasting, vol. 35, no. 4, pp. 1548-1560, September 2019. doi: 10.1016/j.ijforecast.2018.07.006
[3] H. He, M. Sun, X. Li and I. A. Mensah, "A novel crude oil price trend prediction method: Machine learning classification algorithm based on multi-modal data features," Energy, vol. 244, 122706, February 2022. doi: 10.1016/j.energy.2021.122706
[4] N. Jha, H. K. Tanneru, S. Palla and I. H. Mafat, "Multivariate analysis and forecasting of the crude oil prices: Part I–Classical machine learning approaches," Energy, vol. 296, 131185, April 2024. doi: 10.1016/j.energy.2024.131185
[5] J. An, A. Mikhaylov and N. Moiseev, "Oil price predictors: Machine learning approach," International Journal of Energy Economics and Policy, vol. 9, no. 5, pp. 1-6, June 2019. doi: 10.32479/ijeep.7597
[6] S. H. S, "Crude oil price forecasting using machine learning," International Journal of Advanced Scientific Innovation, vol. 1, no. 1, pp. 1-11, December 2020. doi: 10.5281/zenodo.4641697
[7] C. P. Obite, A. Chukwu, D. C. Bartholomew, U. I. Nwosu and G. E. Esiaba, "Classical and machine learning modeling of crude oil production in Nigeria: Identification of an eminent model for application," Energy Reports, vol. 7, 3497-3505, November 2021. doi: 10.1016/j.egyr.2021.06.005
[8] A. Jahandoost, M. Houshmand and S. A. Hosseini, "Prediction of west texas intermediate crude-oil price using hybrid attention-based deep neural networks: a comparative study," in Proceedings of the 13th International Conference on Computer and Knowledge Engineering (ICCKE), 01-02 November 2023, Mashhad, Iran, Islamic Republic of [Online]. Available: IEEE Xplore, https://ieeexplore.ieee.org/abstract/document/10326291?casa_token=pgOS94waRCUAAAAA:pv-FU0oRnJFzzZxotL0Dv01ca5ZKMvira7E-VYSnkeX_qHBG-S5GMenzF2wmyMdB86nyY-NTOw. [Accessed: 29 January 2026]

[9] P. Foroutan and S. Lahmiri, "Deep learning systems for forecasting the prices of crude oil and precious metals," Financial Innovation, vol. 10, no.1, 111, July 2024. doi: 10.1186/s40854-024-00637-z
[10] M. Bi, Z. Liu and X. Yang, "A Hybrid Prediction Model for International Crude Oil Price Based on Variational Mode Decomposition with BiTCN-BiGRU-Attention Deep Learning Techniques," October 2024. Preprint available at Research Square. doi: 10.21203/rs.3.rs-5284460/v1
[11] J. Zeng,W. J. Xie, Z. Y. Wei, Z. L. Wei, H. Y. An, X. W. Jiang, B. Guo and M. Wang, "GBK-Net: A Hybrid GARCH--BiLSTM--KAN Model for Crude Oil Price Forecasting," September 2025. Preprint available at Preprints. doi: 10.20944/preprints202509.0646.v1
[12] M. Türkoğlu and D. Hanbay, "Plant recognition system based on extreme learning machine by using shearlet transform and new geometric features," Journal of the Faculty of Engineering and Architecture of Gazi University, vol. 34, no. 4, pp. 2097-2112, June 2019. doi: 10.17341/gazimmfd.423674
[13] O. Polat, M. Türkoğlu, H. Polat, S. Oyucu, H. Üzen, F. Yardımcı and A. Aksöz, "Multi-stage learning framework using convolutional neural network and decision tree-based classification for detection of DDoS pandemic attacks in SDN-based SCADA systems," Sensors, vol. 24, no. 3, 1040, February 2024. doi: 10.3390/s24031040
[14] A. I. Simsek, E. Koç, B. D. Tasdemir, A. Aksöz, M. Turkoglu, M. and A. Sengur, "Deep Learning Forecasting Model for Market Demand of Electric Vehicles," Applied Sciences, vol. 14, no. 23, 10974, November 2024. doi: 10.3390/app142310974
[15] B. Zhao, H. Lu, S. Chen, J. Liu and D. Wu, "Convolutional neural networks for time series classification," Journal of Systems Engineering and Electronics, vol. 28, no. 1, pp. 162-169, February 2017. doi: 10.21629/JSEE.2017.01.18
[16] E. Soylu, S. Gül, K. A. Koca, M. Türkoğlu, M. Terzi and A. Şengür, "Speech signal-based accurate neurological disorders detection using convolutional neural network and recurrent neural network based deep network," Engineering Applications of Artificial Intelligence, vol. 149, 110558, June 2025. doi: 10.1016/j.engappai.2025.110558
[17] S. Hochreiter and J. Schmidhuber, "Long short-term memory," Neural Computation, vol. 9, no. 8, pp. 1735-1780, November 1997. doi: 10.1162/neco.1997.9.8.1735
[18] M. Turkoglu, "Defective egg detection based on deep features and Bidirectional Long-Short-Term-Memory," Computers and Electronics in Agriculture, vol. 185, 106152, June 2021. doi: 10.1016/j.compag.2021.106152
[19] I. D. Mienye, T. G. Swart and G. Obaido, "Recurrent neural networks: A comprehensive review of architectures, variants, and applications," Information, vol. 15, no. 9, 517, August 2024. doi: 10.3390/info15090517
[20] A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser and I. Polosukhin, "Attention is all you need," Advances in Neural İnformation Processing Systems, vol. 30, December 2017. ISBN: 9781510860964
[21] H. Üzen, M. Turkoglu, D. Ozturk and D. Hanbay, "A novel hybrid attention gate based on vision transformer for the detection of surface defects," Signal, Image and Video Processing, vol. 18, no. 10, pp. 6835-6851, June 2024. doi: 10.1007/s11760-024-03355-2
[22] K. He, X. Zhang, S. Ren and J. Sun, "Deep residual learning for image recognition," in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 27-30 June 2016, Las Vegas, NV, USA [Online]. Available: IEEE, https://www.computer.org/csdl/proceedings/cvpr/2016/12OmNqH9hnp. [Accessed: 12 February 2026]
[23] Ö. Dürrani, S. İşgüzar, A. İmak, T. Ateşşahin, Z. Cömert, S. Z. Dürrani and M. Türkoğlu, "MobileNet-and attention-based MLP-Mixer architecture for geographical-region recognition of a marine fish (Perciformes: Carangidae) using otolith images," Expert Systems with Applications, vol. 291, 128585, October 2025. doi: 10.1016/j.eswa.2025.128585

Details

Primary Language

English

Subjects

Modelling and Simulation, Artificial Intelligence (Other)

Journal Section

Research Article

Authors

Onur Polat
0000-0001-9313-4910
Türkiye

Neslihan Parlak
0009-0006-4248-9504
Türkiye

Esra Söğüt ^*
0000-0002-0051-2271
Türkiye

Eda Fendoğlu
0000-0003-4092-7137
Türkiye

Muammer Türkoğlu
0000-0002-2377-4979
Türkiye

Publication Date

August 31, 2026

Submission Date

April 22, 2026

Acceptance Date

June 17, 2026

Published in Issue

Year 2026 Volume: 12 Number: 2

IZ

https://izlik.org/JA65JL48UE

Cite

RIS / Bibtex

APA

Polat, O., Parlak, N., Söğüt, E., Fendoğlu, E., & Türkoğlu, M. (2026). Explainable and Robust Oil Price Forecasting Using Hybrid Deep Learning and SHAP Interpretability: A Long-Term Analysis of WTI Prices from 1986 to 2025. Gazi Journal of Engineering Sciences, 12(2), 201-220. https://izlik.org/JA65JL48UE

AMA

1.Polat O, Parlak N, Söğüt E, Fendoğlu E, Türkoğlu M. Explainable and Robust Oil Price Forecasting Using Hybrid Deep Learning and SHAP Interpretability: A Long-Term Analysis of WTI Prices from 1986 to 2025. GJES. 2026;12(2):201-220. https://izlik.org/JA65JL48UE

Chicago

Polat, Onur, Neslihan Parlak, Esra Söğüt, Eda Fendoğlu, and Muammer Türkoğlu. 2026. “Explainable and Robust Oil Price Forecasting Using Hybrid Deep Learning and SHAP Interpretability: A Long-Term Analysis of WTI Prices from 1986 to 2025”. Gazi Journal of Engineering Sciences 12 (2): 201-20. https://izlik.org/JA65JL48UE.

EndNote

Polat O, Parlak N, Söğüt E, Fendoğlu E, Türkoğlu M (August 1, 2026) Explainable and Robust Oil Price Forecasting Using Hybrid Deep Learning and SHAP Interpretability: A Long-Term Analysis of WTI Prices from 1986 to 2025. Gazi Journal of Engineering Sciences 12 2 201–220.

IEEE

[1]O. Polat, N. Parlak, E. Söğüt, E. Fendoğlu, and M. Türkoğlu, “Explainable and Robust Oil Price Forecasting Using Hybrid Deep Learning and SHAP Interpretability: A Long-Term Analysis of WTI Prices from 1986 to 2025”, GJES, vol. 12, no. 2, pp. 201–220, Aug. 2026, [Online]. Available: https://izlik.org/JA65JL48UE

ISNAD

Polat, Onur - Parlak, Neslihan - Söğüt, Esra - Fendoğlu, Eda - Türkoğlu, Muammer. “Explainable and Robust Oil Price Forecasting Using Hybrid Deep Learning and SHAP Interpretability: A Long-Term Analysis of WTI Prices from 1986 to 2025”. Gazi Journal of Engineering Sciences 12/2 (August 1, 2026): 201-220. https://izlik.org/JA65JL48UE.

JAMA

1.Polat O, Parlak N, Söğüt E, Fendoğlu E, Türkoğlu M. Explainable and Robust Oil Price Forecasting Using Hybrid Deep Learning and SHAP Interpretability: A Long-Term Analysis of WTI Prices from 1986 to 2025. GJES. 2026;12:201–220.

MLA

Polat, Onur, et al. “Explainable and Robust Oil Price Forecasting Using Hybrid Deep Learning and SHAP Interpretability: A Long-Term Analysis of WTI Prices from 1986 to 2025”. Gazi Journal of Engineering Sciences, vol. 12, no. 2, Aug. 2026, pp. 201-20, https://izlik.org/JA65JL48UE.

Vancouver

1.Onur Polat, Neslihan Parlak, Esra Söğüt, Eda Fendoğlu, Muammer Türkoğlu. Explainable and Robust Oil Price Forecasting Using Hybrid Deep Learning and SHAP Interpretability: A Long-Term Analysis of WTI Prices from 1986 to 2025. GJES [Internet]. 2026 Aug. 1;12(2):201-20. Available from: https://izlik.org/JA65JL48UE