Demand Forecasting for Cross-border Retail Commodities in the New Retail Era: A Hybrid Model Integrating Temporal Granularity and Categorical Dynamics

Through the lens of the Temporal-Synergic Cross-Correlation (TSCC) framework, this study fills important gaps in cross-border demand forecasting. Overall, analyzing 9,754 transactions (2022–2024) our approach captures multi-scale temporal decomposition synergized with SHAP-based feature engineering and outperforms the LSTM-RF hybrids by 22% precision and discovers 14.7% YoY revenue growth. Notably, the innovations tackled temporal-categorical disconnects and 92% holiday prediction precision — 18.4% on global metrics. The framework bridges theoretical models and operational logistics, harvesting trials and errors to overcome cross-border commerce lead-time preserves.