Research on World Agricultural Economy

Forecasting Agricultural Trade Based on TCN-LightGBM Models: A Data-Driven Decision

Tianwen Zhao

Department of Trade and Logistics, Daegu Catholic University, Gyeongsan 38430, Republic of Korea

Guoqing Chen

Mathematical Modeling Research Center, Chengdu Jincheng College, Chengdu 611731, China; Department of Mathematics, Faculty of Science, Mahasarakham University, Maha Sarakham 44150, Thailand

Thom Gatewongsa

The Research Institute of Northeastern Art and Culture, Mahasarakham University, Maha Sarakham 44150, Thailand

Piyapatr Busababodhin

Department of Mathematics, Faculty of Science, Mahasarakham University, Maha Sarakham 44150, Thailand; The Research Institute of Northeastern Art and Culture, Mahasarakham University, Maha Sarakham 44150, Thailand

DOI: https://doi.org/10.36956/rwae.v6i1.1429

Received: 28 October 2024 | Revised: 13 November 2024 | Accepted: 14 November 2024 | Published Online: 16 January 2025

---

Abstract

Facing the increasing complexity and dynamic fluctuations of the global agricultural trade market, accurate forecasting plays a key role in supporting agricultural policy formulation, stabilising the market and optimising resource allocation. In order to increase the precision and stability of agricultural trade predictions, this research suggests a hybrid model built on a temporal convolution network (TCN) and a lightweight gradient boosting tree (LightGBM). The TCN module effectively captures the long-term dependence characteristics of time series data through dilated convolution, which improves the model’s ability to identify seasonal and periodic trends. The LightGBM module, on the other hand, makes use of the characteristics of gradient boosting decision trees and excels at efficiently handling nonlinear relationships and avoiding overfitting. Experimental results show that the TCN-LightGBM model outperforms traditional models in terms of mean square error (MSE), mean absolute error (MAE) and prediction accuracy. Specifically, compared with ARIMA, LSTM, TCN alone or LightGBM alone, TCN-LightGBM achieves a prediction accuracy of 91.3% on the test data, with MSE and MAE of 0.021 and 0.115 respectively, significantly improving prediction accuracy and stability. In addition, parameter sensitivity analysis shows that the TCN-LightGBM model maintains a highly consistent prediction trend under different parameter configurations, which verifies the robustness of the model and its practical application value. This study provides a data-driven decision support tool with high accuracy and strong stability, providing a new solution for agricultural trade forecasting and other complex time series prediction tasks.

Keywords: Agricultural Trade Forecasting; TCN; LightGBM; Data-Driven; Decision Support

---

References

[1] Wang, Y., Chen, J., Chen, X., et al., 2020. Short-term load forecasting for industrial customers based on TCN-LightGBM. IEEE Transactions on Power Systems. 36(3), 1984–1997. DOI: https://doi.org/10.1109/TPWRS.2020.3028133

[2] Nguyen, K.-T., Tran, T.-N., Nguyen, H.-T., 2024. Research on the influence of hyperparameters on the LightGBM model in load forecasting. Engineering, Technology & Applied Science Research. 14(5), 17005–17010. DOI: https://doi.org/10.48084/etasr.8266

[3] Ma, L., Cai, J., Dai, X., et al., 2022. Research on rockburst risk level prediction method based on LightGBM−TCN−RF. Applied Sciences. 12(16), 8226. DOI: https://doi.org/10.3390/app12168226

[4] Yue, Y., 2023. Integrated forecasting model based on LSTM and TCN for short-term power load forecasting. Advances in Engineering Technology Research. 6(1), 540. DOI: https://doi.org/10.56028/aetr.6.1.540.2023

[5] Heng, L., Hao, C., Nan, L., 2024. Load forecasting method based on CEEMDAN and TCN-LSTM. PLOS ONE. 19(7), e0300496. DOI: https://doi.org/10.1371/journal.pone.0300496

[6] Iyiola, K., Alzubi, A., 2024. Out-of-stock prediction model using Buzzard Coney Hawk Optimization-based LightGBM-enabled deep temporal convolutional neural network. Applied Sciences. 14(13), 5906. DOI: https://doi.org/10.3390/app14135906

[7] Li, M., Zhang, T., Yang, H., et al., 2024. Multiple load forecasting of integrated renewable energy system based on TCN-FECAM-Informer. Energies. 17(20), 5181. DOI: https://doi.org/10.3390/en17205181

[8] Huang, J., Zhang, X., Jiang, X., 2023. Short-term power load forecasting based on the CEEMDAN-TCN-ESN model. PLOS ONE. 18(10), e0284604. DOI: https://doi.org/10.1371/journal.pone.0284604

[9] Xiang, X., Yuan, T., Cao, G., et al., 2024. Short-term electric load forecasting based on signal decomposition and improved tcn algorithm. Energies. 17(8), 1815. DOI: https://doi.org/10.3390/en17081815

[10] Liu, M., Qin, H., Cao, R., et al., 2022. Short-term load forecasting based on improved TCN and DenseNet. IEEE Access. 10, 115945–115957. DOI: https://doi.org/10.1109/ACCESS.2022.3218374

[11] Samal, K.Krishna., 2024. Auto imputation enabled deep Temporal Convolutional Network (TCN) model for pm2.5 forecasting. ICST Transactions on Scalable Information Systems. 11, 1–15. DOI: https://doi.org/10.4108/eetsis.5102

[12] Zhang, R., Sun, F., Song, Z., et al., 2021. Short-term traffic flow forecasting model based on GA-TCN. Journal of Advanced Transportation. 2021(722), 1–13. DOI: https://doi.org/10.1155/2021/1338607

[13] Zan, S., Zhang, Q., 2023. Short-term power load forecasting based on an EPT-VMD-TCN-TPA model. Applied Sciences. 13(7), 4462. DOI: https://doi.org/10.3390/app13074462

[14] Han, H., Peng, J., Ma, J., et al., 2024. A short-term electricity load forecasting model: CEEMDAN-SE-VMD+SelfAttention-TCN Fusion model. DOI: https://doi.org/10.21203/rs.3.rs-4881138/v1

[15] Zheng, Q., Zheng, J., Mei, F., et al., 2023. TCN-GAT multivariate load forecasting model based on SHAP value selection strategy in integrated energy system. Frontiers in Energy Research. 11, 1208502. DOI: https://doi.org/10.3389/fenrg.2023.1208502

[16] Xiang, X., Li, X., Zhang, Y., et al., 2024. A short-term forecasting method for photovoltaic power generation based on the TCN-ECANet-GRU hybrid model. Scientific Reports. 14(1), 6744. DOI: https://doi.org/10.1038/s41598-024-56751-6

[17] Saepudin, D., Rabbani, E., Navialdy, D., et al., 2024. Water level time series forecasting using TCN study case in Surabaya. Jurnal Online Informatika. 9(1), 61–69. DOI: https://doi.org/10.15575/join.v9i1.1312

[18] Zuo, C., Wang, J., Liu, M., et al., 2023. An ensemble framework for short-term load forecasting based on TimesNet and TCN. Energies. 16(14), 5330. DOI: https://doi.org/10.3390/en16145330

[19] Hartanto, A., Kholik, Y., Pristyanto, Y., 2023. Stock price time series data forecasting using the Light Gradient Boosting Machine (LightGBM) model. JOIV: International Journal on Informatics Visualization. 7(4), 2270. DOI: https://doi.org/10.30630/joiv.7.4.01740

[20] Xiong, Q., Liu, M., Li, Y., et al., 2023. Short-term load forecasting based on VMD and deep TCN-based hybrid model with self-attention mechanism. Applied Sciences. 13(22), 12479. DOI: https://doi.org/10.3390/app132212479

[21] Liu, M., Sun, X., Wang, Q., et al., 2022. Short-term load forecasting using EMD with feature selection and TCN-based deep learning model. Energies. 15(19). 7170. DOI: https://doi.org/10.3390/en15197170

[22] Liu, Y., Wang, Y., Liu, X., et al., 2024. Research on runoff prediction based on Time2Vec-TCN-Transformer driven by multi-source data. Electronics. 13(14), 2681. DOI: https://doi.org/10.3390/electronics13142681

[23] Simaiya, S., Dahiya, M., Tomar, S., et al., 2024. A transfer learning-based hybrid model with LightGBM for smart grid short-term energy load prediction. Energy Exploration & Exploitation. 13(22), 01445987241256472. DOI: https://doi.org/10.1177/01445987241256472

[24] Li, Q., Wang, D., Liu, F., et al., 2024. LightGBM hybrid model based DEM correction for forested areas. PLOS ONE. 19(10), e0309025. DOI: https://doi.org/10.1371/journal.pone.0309025

[25] Limouni, T., Yaagoubi, R., Bouziane, K., et al., 2023. Accurate one step and multistep forecasting of very short-term PV power using LSTM-TCN model. Renewable Energy. 205, 1010–1024. DOI: https://doi.org/10.1016/j.renene.2023.01.118

[26] Kriuchkova, A., Toloknova, V., Drin, S., 2024. Predictive model for a product without history using LightGBM. Pricing model for a new product. Mohyla Mathematical Journal. 6, 6–13. DOI: https://doi.org/10.18523/2617-7080620236-13 (In Russian)

[27] Hamiane, S., Ghanou, Y., Khalifi, H., et al., 2024. Comparative analysis of LSTM, ARIMA, and hybrid models for forecasting future GDP. International Information and Engineering Technology Association. 29(3), 853–861. DOI: https://doi.org/10.18280/isi.290306