International Journal of Multidisciplinary Engineering Sciences | Volume 1 Issue 2 | Pages: 1-12 | Doi : 10.67914/ijmes/rsa/1.2/1-12

Research Article

OPEN ACCESS | Published on : 27-Aug-2026

Machine Learning-Based Sales Forecasting for Retail Supply Chain Optimization

Abatcha Alhaji Kurna
• Department of Computer Science, Federal Polytechnic Kaltungo, Gombe State, Nigeria.

Ahmad Tukur
• Department of Information and Communication Technology, Federal Polytechnic Kaltungo, Gombe State, Nigeria.


• Aminu Adamu Ahmed
• Department of Information and Communication Technology, Federal Polytechnic Kaltungo, Gombe State, Nigeria.

Abstract

Accurate, forward-looking sales forecasts are a prerequisite for effective retail supply chain optimization, informing procurement, warehousing, and replenishment decisions well before demand materialises. This study develops a lag-feature-based machine learning forecasting framework for monthly, category-level retail revenue, using a transactional e-commerce dataset of 5,000 orders spanning four product categories over a multi-year horizon. Historical revenue, quantity, discount, and order-count values were aggregated into a category-month panel (647 usable observations after lag construction) and enriched with first-, second-, and third-order revenue lags and a three-month rolling mean, following established lag-feature forecasting practice. Six forecasting approaches were compared on a chronologically held-out test period: a naive persistence baseline, a three-month moving-average baseline, Linear Regression, Random Forest, Gradient Boosted Trees, and a windowed Artificial Neural Network (Multilayer Perceptron). Because the specialised recurrent deep-learning architectures originally specified for this study: Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Bidirectional GRU (Bi-GRU) networks require TensorFlow or PyTorch, and these libraries (together with internet access to install them) were unavailable in the computational environment used for this research, the windowed MLP is reported explicitly as a substitute deep-learning benchmark rather than as a recurrent architecture, and this substitution is discussed in detail as a limitation. Linear Regression achieved the strongest held-out performance (RMSE = $1,762, R² = 0.831, MAPE = 23.96%), narrowly ahead of Random Forest (R² = 0.811) and Gradient Boosted Trees (R² = 0.794), while both baselines and the windowed MLP underperformed the regression-based learners. Feature-importance analysis showed that order quantity, rather than the revenue lags themselves, was overwhelmingly the dominant predictor (90.7% relative importance) a finding attributable to the near-deterministic relationship between quantity, unit price, and revenue in this dataset. The study concludes that classical regression and ensemble methods, when combined with carefully engineered lag and rolling-window features, provide an accurate and computationally inexpensive foundation for supply chain demand forecasting, and outlines a concrete roadmap including LSTM/GRU/Bi-GRU implementation once suitable deep-learning infrastructure is available for extending this framework toward sequence-aware, uncertainty-calibrated forecasting.

Keywords

Sales Forecasting, Supply Chain Optimization, Lag Features, Machine Learning, Time-Series Regression, Retail Demand

References

• Ben Elmir, W., Hemmak, A., & Senouci, B. (2023). Smart platform for data blood bank management: Forecasting demand in blood supply chain using machine learning. Information, 14(1), 31. https://doi.org/10.3390/info14010031

  Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. https://doi.org/10.1023/A:1010933404324

  Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 785–794). ACM. https://doi.org/10.1145/2939672.2939785

  Douaioui, K., Oucheikh, R., Benmoussa, O., & Mabrouki, C. (2024). Machine learning and deep learning models for demand forecasting in supply chain management: A critical review. Applied System Innovation, 7(5), 93. https://doi.org/10.3390/asi7050093

  Falatouri, T., Darbanian, F., Brandtner, P., & Udokwu, C. (2022). Predictive analytics for demand forecasting—a comparison of SARIMA and LSTM in retail SCM. Procedia Computer Science, 200, 993–1003. https://doi.org/10.1016/j.procs.2022.01.298

  Friedman, J. H. (2001). Greedy function approximation: A gradient boosting machine. The Annals of Statistics, 29(5), 1189–1232.

  Hastie, T., Tibshirani, R., & Friedman, J. (2009). The elements of statistical learning: Data mining, inference, and prediction (2nd ed.). Springer.

  Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8), 1735–1780. https://doi.org/10.1162/neco.1997.9.8

  Huber, J., & Stuckenschmidt, H. (2020). Daily retail demand forecasting using machine learning with emphasis on calendric special days. International Journal of Forecasting, 36(4), 1420–1438. https://doi.org/10.1016/j.ijforecast.2020.02.005

  Ji, S., Wang, X., Zhao, W., & Guo, D. (2019). An application of a three-stage XGBoost-based model to sales forecasting of a cross-border e-commerce enterprise. Mathematical Problems in Engineering, 2019, Article 8503252. https://doi.org/10.1155/2019/8503252

  Lu, J., Zheng, X., Nervino, E., Li, Y., Xu, Z., & Xu, Y. (2024). Retail store location screening: A machine learning-based approach. Journal of Retailing and Consumer Services, 77, Article 103620. https://doi.org/10.1016/j.jretconser.2023.103620

  Ma, S., & Fildes, R. (2021). Retail sales forecasting with meta-learning. European Journal of Operational Research, 288(1), 111–128. https://doi.org/10.1016/j.ejor.2020.05.038

  Mustapha, O. O., & Sithole, T. (2025). Forecasting retail sales using machine learning models. American Journal of Statistics and Actuarial Science. https://ajpojournals.org/journals/AJSAS/article/view/2679

  Oncioiu, I., Bunget, O. C., Türkes, M. C., Capusneanu, S., Topor, D. I., Tamas, A. S., Rakos, I. S., & Hint, M. S. (2019). The impact of big data analytics on company performance in supply chain management. Sustainability, 11(18), 4864. https://doi.org/10.3390/su11184864

  Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, R., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., & Duchesnay, É. (2011). Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12, 2825–2830.

  Petropoulos, F., Apiletti, D., Assimakopoulos, V., Babai, M. Z., Barrow, D. K., Ben Taieb, S., Bergmeir, C., Bessa, R. J., Bijak, J., Boylan, J. E., et al. (2022). Forecasting: Theory and practice. International Journal of Forecasting, 38(3), 705–871. https://doi.org/10.1016/j.ijforecast.2021.11.001

  Phyu, M. M., & Khine, M. T. (2023). Retail demand forecasting using sequence to sequence long short-term memory networks. In Proceedings of the 2023 IEEE Conference on Computer Applications (ICCA) (pp. 208–213). IEEE.

  Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1986). Learning representations by back-propagating errors. Nature, 323(6088), 533–536. https://doi.org/10.1038/323533a0

  ToolsGroup. (2026). Machine learning in demand planning: How to boost forecasting. https://www.toolsgroup.com/blog/machine-learning-in-demand-planning-how-to-boost-forecasting/

  Vukovic, D. B., Spitsina, L., Gribanova, E., Spitsin, V., & Lyzin, I. (2023). Predicting the performance of retail market firms: Regression and machine learning methods. Mathematics, 11(8), 1916. https://doi.org/10.3390/math11081916

  Weber, F., & Schütte, R. (2019). A domain-oriented analysis of the impact of machine learning—the case of retailing. Big Data and Cognitive Computing, 3(1), 11. https://doi.org/10.3390/bdcc3010011

  Xie, L., Liu, J., & Wang, W. (2024). Predicting sales and cross-border e-commerce supply chain management using artificial neural networks and the Capuchin search algorithm. Scientific Reports, 14, 13340. https://doi.org/10.1038/s41598-024-62368-6