Hybrid Forecasting Systems for Inventory Optimization Using Prophet and Reinforcement Learning
DOI:
https://doi.org/10.22399/ijcesen.4123Keywords:
Inventory Optimization, Hybrid Forecasting, Time Series Prediction, Reinforcement Learning, Demand Forecasting, Supply Chain ManagementAbstract
Effective inventory supervision aids modern supply chain operations. However, established methods of forecasting often falter when faced with ever-changing demand, multiple seasons, and promotional peaks, resulting in either stock surplus or shortage, both of which are costly. This study introduces a forecasting methodology that combines Facebook Prophet, a time-series forecasting tool, and Reinforcement Learning (RL) for inventory optimisation. Prophet analyses and forecasts demands by recognising the intricate temporal attributes in the historical sales data, while RL uses these forecasts to continuously refine ordering policies, inventory holding, and shortage costs. The experiments on retail datasets confirm that the new system decreases the forecasting errors by 25% when compared with ARIMA and LSTM and improves inventory service levels by 15–20%, all while cutting the overall inventory expenses. These results underscore the significance of merging statistical forecasting and intelligent decision-making and provide a utilitarian methodology for supply chains to tackle demand and operational variability.
References
[1] Ligentia. (2024). Overcoming inventory management hurdles: Transforming supply chains for efficiency and profitability. Retrieved from
[2] Gao, Y. (2024). A comparative study of ARIMA and ETS models for time series forecasting. ResearchGate.
[3] Hyndman, R. J., & Athanasopoulos, G. (2018). Forecasting: principles and practice. 2nd ed. Retrieved from
[4] Gijsbrechts, J., Carias, C., & van Woensel, T. (2018). Deep reinforcement learning for inventory optimization: A comparative analysis. Computers & Operations Research, 100, 311–324.
[5] Doborjginidze, G., et al. (2021). Optimization of Inventory Management in the Supply Chain. ResearchGate.
[6] Fildes, R., & Goodwin, P. (2017). Demand Forecasting: Evidence-based methods and their use. Journal of the Operational Research Society, 68(3), 335–345.
[7] Setyadi, H. A., Amin, B. A., & Widodo, P. (2024). Implementation of Economic Order Quantity and Reorder Point methods in inventory management information systems. Journal of Information Systems and Informatics, 6(1), 1–10.
[8] Raban, M., & Gordon, A. (2024). Sales Forecasting Models: Comparison between ARIMA, LSTM and Prophet. Journal of Computer Science and Software Engineering, 12(3), 1222–1230.
[9] Zhao, Y., et al. (2021). Deep reinforcement learning for inventory control: A roadmap. Computers & Operations Research, 129, 105198.
[10] Chopra, S., & Meindl, P. (2020). Supply Chain Management: Strategy, Planning & Operation. 7th Edition. Pearson.
[11] Box, G. E. P., Jenkins, G. M., & Reinsel, G. C. (2015). Time Series Analysis: Forecasting and Control. 5th Edition. Wiley.
[12] Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785–794.
[13] Li, X., et al. (2020). Hybrid approaches for demand forecasting and inventory optimization: A review. European Journal of Operational Research, 286(2), 405–423.
[14] Taylor, S., & Letham, B. (2018). Forecasting at Scale. The American Statistician, 72(1), 37–45.
[15] Hochreiter, S., & Schmidhuber, J. (1997). Long Short-Term Memory. Neural Computation, 9(8), 1735–1780.
[16] Li, X., et al. (2020). Reinforcement learning for inventory management in e-commerce. European Journal of Operational Research, 285(3), 1057–1073.
[17] Zhang, Y., et al. (2019). Integrating Prophet with inventory heuristics for retail demand forecasting. Computers & Industrial Engineering, 135, 98–110.
[18] Kumar, R., & Sharma, P. (2021). Deep Q-Learning for inventory optimization in FMCG supply chains. International Journal of Production Research, 59(24), 7472–7488.
[19] Ahmed, S., et al. (2022). Hybrid LSTM and reinforcement learning for manufacturing inventory control. Computers & Industrial Engineering, 167, 107950.
[20] Wang, L., & Chen, H. (2021). Integrating ARIMA and reinforcement learning for dynamic retail inventory management. Expert Systems with Applications, 171, 114568.
[21] Ghosh, S., et al. (2020). Combining Prophet and genetic algorithms for inventory optimization. Computers & Operations Research, 118, 104949.
[22] Singh, A., & Verma, R. (2022). Multi-agent reinforcement learning for coordination in multi-echelon supply chains. Journal of Manufacturing Systems, 64, 290–303.
[23] Kiran, R., & Singh, P. (2021). Data quality considerations for machine learning in supply chain management. Computers & Industrial Engineering, 153, 107096.
[24] Chen, X., et al. (2020). Retail sales datasets for demand forecasting research: Characteristics and preprocessing. Expert Systems with Applications, 149, 113254.
[25] Fildes, R., Goodwin, P., & Lawrence, M. (2019). Features affecting demand forecasting accuracy in retail. International Journal of Forecasting, 35(2), 627–638.
[26] Schulman, J., et al. (2017). Proximal Policy Optimization Algorithms. arXiv preprint arXiv:1707.06347. https://arxiv.org/abs/1707.06347
[27] Konda, V., & Tsitsiklis, J. (2000). Actor-Critic Algorithms. SIAM Journal on Control and Optimization, 42(4), 1143–1166.
[28] Feurer, M., et al. (2019). Auto-sklearn 2.0: The next generation. Automated Machine Learning, 202–220.
[29] Ivanov, D., et al. (2021). Sustainable inventory management: integrating environmental metrics in supply chain optimization. Journal of Cleaner Production, 279, 123456.
[30] Chopra, S., Meindl, P., & Kalra, D. (2021). Modern trends in supply chain risk management. International Journal of Production Economics, 232, 107-121.
[31] He, H., & Garcia, E. A. (2009). Learning from imbalanced data. IEEE Transactions on Knowledge and Data Engineering, 21(9), 1263–1284.
[32] Pan, S. J., & Yang, Q. (2010). A survey on transfer learning. IEEE Transactions on Knowledge and Data Engineering, 22(10), 1345–1359.
[33] Petropoulos, F., Makridakis, S., & Assimakopoulos, V. (2021). Forecasting in supply chains: The state of the art. International Journal of Forecasting, 37(3), 1101–1123.
[34] Li, X., et al. (2021). Reinforcement learning for dynamic pricing and inventory management. European Journal of Operational Research, 291(1), 1–14.
[35] Silver, D., et al. (2016). Mastering the game of Go with deep neural networks and tree search. Nature, 529, 484–489.
[36] Bottou, L., et al. (2018). Optimization methods for large-scale machine learning. SIAM Review, 60(2), 223–311.
[37] Li, X., & Zhao, Y. (2022). Inventory optimization under stochastic demand using reinforcement learning. Computers & Industrial Engineering, 164, 107855.
[38] Lim, M. K., & Lu, L. (2019). Integrating machine learning and optimization for supply chain demand forecasting. European Journal of Operational Research, 274(2), 667–682.
[39] Van Roy, B., et al. (2006). Reinforcement learning in continuous time and space. In Reinforcement Learning, Springer, 213–242.
[40] Rajeswaran, A., et al. (2017). EPOpt: Learning robust neural network policies using model ensembles. arXiv preprint arXiv:1610.01283.
[41] Chen, W., et al. (2022). Combining deep learning and optimization for multi-echelon inventory control. Computers & Industrial Engineering, 169, 108253.
[42] Guo, S., et al. (2020). Inventory management under lead-time uncertainty: A machine learning approach. International Journal of Production Research, 58(23), 7057–7075.
[43] Choi, T. M., et al. (2018). Big data analytics in operations management. Production and Operations Management, 27(10), 1868–1882. https://doi.org/10.1111/poms.12850
[44] Zhang, H., et al. (2021). Deep reinforcement learning for inventory replenishment under stochastic demand. Expert Systems with Applications, 174, 114800.
[45] Tang, O., & Musa, S. N. (2011). Identifying risk issues and research advancements in supply chain risk management. International Journal of Production Economics, 133(1), 25–34.
[46] Wang, X., et al. (2020). Demand forecasting with temporal convolutional networks. International Journal of Forecasting, 36(4), 1402–1417.
[47] Li, X., et al. (2021). Multi-agent reinforcement learning for decentralized inventory management. Computers & Industrial Engineering, 156, 107220.
[48] Gaur, V., et al. (2020). Machine learning approaches in supply chain demand forecasting: A review. Computers & Industrial Engineering, 149, 106832.
[49] Benjaafar, S., et al. (2018). Operations management in the age of big data. Production and Operations Management, 27(10), 1812–1825.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 International Journal of Computational and Experimental Science and Engineering
This work is licensed under a Creative Commons Attribution 4.0 International License.
