Stock Inventory Prediction Model Based on SSA-VMD-XGBoost

Accurate stock inventory prediction is crucial for supply chain management. However, traditional methods struggle to handle complex time series patterns effectively, which adversely impacts forecasting accuracy. To address this issue, this paper proposes an SSA-VMD-XGBoost model to enhance the accuracy and generalization capability of inventory prediction. Firstly, Variational Mode Decomposition (VMD) is employed to decompose the original inventory data into multiple sub-sequences, extracting key features while reducing noise interference. Then, the Sparrow Search Algorithm (SSA) is used to optimize the VMD decomposition parameters, improving the quality of signal decomposition. Finally, the optimized features are fed into XGBoost for training and prediction, enhancing the model's learning capability. Experimental results demonstrate that, compared to XGBoost and VMD+XGBoost, the proposed SSA-VMD-XGBoost model achieves superior predictive performance on the training, validation, and test sets, with an <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{R}^{\mathbf{2}}$</tex> score of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{0. 9 5 5 1 4}$</tex> on the test set, significantly improving prediction accuracy and model stability. By integrating VMD for data decomposition and SSA for hyperparameter optimization, this method substantially enhances the predictive capability of XGBoost. The proposed approach not only improves time-series forecasting accuracy but also strengthens model stability and generalization ability, providing an effective solution for high-precision forecasting tasks.