Optimising Inventory Management Strategies for Cost Reduction in Supply Chains: A Systematic Review

Abstract

This research proposes a data-driven inventory optimization methodology integrating advanced demand forecasting with traditional inventory models to reduce costs and improve service levels. Using historical sales data, seasonal patterns, and external demand drivers, machine learning models (e.g., Random Forest) are developed to predict future demand more accurately than classical forecasting techniques. The forecast output feeds into the Economic Order Quantity (EOQ) model and refined safety stock calculations, adjusting reorder points dynamically to minimize costs related to overstock and stockouts. Scenario analyses and sensitivity testing demonstrate robustness under varying demand conditions, showing improved inventory efficiency and reduced total cost of ownership. The study highlights the applicability of combining AI-based forecasting with established inventory control techniques to build resilient and cost-effective supply chains.