Predicting User Purchase Behavior Based on SSA-XGBoost Model

In order to accurately predict user purchasing behavior and help merchants maximize the sales conversion rate and user satisfaction on e-commerce platforms, this paper proposes a user purchasing behavior prediction model based on the combination of the Sparrow Search Algorithm (SSA) and Extreme Gradient Boosting (XGBoost) - the SSA-XGBoost model. This study involves data preprocessing, feature construction, building the XGBoost prediction model, and using the SSA to optimize the parameters of the XGBoost model to ensure the model's performance reaches its optimal state. The findings indicate that the SSA-optimized XGBoost model outperforms the standalone XGBoost model across various evaluation metrics. Specifically, the SSA-XGBoost model's prediction accuracy, F1 score, and AUC are improved compared to the single XGBoost model. These improvements indicate that the SSA-XGBoost model has higher reliability and accuracy in predicting user purchasing behavior, providing merchants with more precise marketing strategies and more efficient operational decision support, thereby enhancing the overall performance and user satisfaction of the e-commerce platform.