Warehouse Demand Variation Prediction and Optimization Based on Machine Learning Algorithms

With the intensification of market competition and the diversification of customer demands, efficient inventory management has become a key factor for the success of enterprises. Warehouse demand forecasting, as a core aspect of inventory management, is of great significance for enterprises to reasonably arrange production, optimize inventory levels, and reduce costs. This paper delves into the prediction and optimization methods of warehouse demand changes based on machine learning algorithms, with a particular focus on three powerful algorithms: Random Forest, XGBoost, and LightGBM.the LightGBM model performed the best. The lowest MSE can provide a lower prediction error, and the highest R² value, which is close to 0.786, can better fit the data. The research results fully demonstrate that machine learning algorithms have significant application value and potential in the field of predicting changes in warehouse demand.