Abstract:

Effective inventory management is essential for reducing costs, minimizing waste, and maintaining product availability. Inventory control techniques traditionally use static forecast models, which cause stockouts or overstocking. We examine the utilization of machine learning (ML) methods for demand forecasting and intelligent replenishment in this paper. We compare several ML algorithms, such as time-series forecasting (ARIMA, LSTM) and regression-based techniques (Random Forest, XGBoost), for improving the accuracy of inventory. Our findings indicate that ML-driven forecasting significantly improves forecast accuracy, optimizes inventories, and enhances supply chain efficiency. Our suggested approach brings together real-time data processing as well as adaptive replenishment schemes, resulting in a more robust and cost-effective inventory system.