Inventory Optimization Using Data Science Technologies for Supply Chain 4.0
Emna Bouazizi,Ayman Khedr,Sherin Elfaioumy,Mohamed Belal
DOI: https://doi.org/10.34028/iajit/21/6/4
2024-01-01
The International Arab Journal of Information Technology
Abstract:In the context of supply chain 4.0, Data Science (DS) can improve operations and inventory management by using statistical and machine learning techniques. Additionally, big data can reveal valuable insights for predictive and prescriptive analytics, which can aid in enhancing competitiveness in today’s business environment. Accurate demand forecasting in supply chain inventory is a fundamental step to improve inventory management, address ordering uncertainties, and minimize costs while meeting customer demands. This requires the efficient use of demand forecasting tools, assessing predictive data analysis techniques. Challenges can arise from inefficient ordering models, selecting accurate forecasting models, and considering issues like over-and underestimation leading to food waste and profit margin impacts, particularly for products with short shelf life. Other challenges include enhancing inventory optimization efficiency, adapting to dynamic demands, and formulating optimal inventory decisions. In this work, we introduce a supply chain 4.0 inventory management approach, where we combined DS techniques, predictive analytics, and big data approach to enhance inventory control. A prediction model is also introduced in order to forecast incoming and outgoing inventory. This model is based on a detailed dataset that takes into account the districts and seasons data. The model’s performance was evaluated based on performance measurements, such as Mean Squared Error (MSE), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE). Significantly, the results show that the random forest algorithm was the best in predicting OUT inventory quantities with an average error equal to 0.011, while the linear regression algorithm produced the best performance in predicting IN inventory quantities with an average error of 0.03
computer science, information systems, artificial intelligence,engineering, electrical & electronic