Freight Value Prediction Model

In the bustling world of online retail, mastering shipping logistics stands out as a vital element for boosting customer loyalty and trimming operational overheads. Our research dives into a machine learning strategy to forecast shipping expenses and arrival times, drawing from historical e-commerce records. We selected XGBoost for its knack in managing intricate datasets, incorporating elements such as shipment methods, package weights, customer engagements, and promotional discounts. This framework empowers logistics managers to refine routes and control budgets more astutely, ultimately elevating service quality and business performance. When tested on an openly accessible dataset, the model delivered impressive outcomes: an RMSE of 12.45 and R² of 0.92 for cost estimations, alongside an accuracy of 0.85 and ROC-AUC of 0.88 for punctuality classifications.