Logistics Demand Fluctuation Forecasting and Real-Time Decision Support Based on XGBoost

In this paper, XGBoost is used to forecast logistics warehousing. We fit the demand of 5 warehouses from 2010 to 2020, and forecast the demand from 2020 to 2025. First, the predictive performance of the model is quantified by a number of evaluation measures, including mean error (ME), root mean square error (RMSE), mean absolute error (MAE), mean percentage error (MPE), mean absolute percentage error (MAPE), and mean absolute normalization error (MASE). These indicators give us a comprehensive perspective on the predictive power of our models. The results show that the model can not only accurately model the historical demand data, but also provide real-time forecast results in the face of future demand fluctuations, and the decision support.