Research and Prediction Based on EWM-TOPSIS and SARIMAX-XGBoost Model

Based on the DPI Dataset, by integrating multidimensional data and utilizing the entropy weight method and TOPSIS to construct an evaluation model, Grey Relational Analysis was used to assess the correlation between the number of subjects and various features. The analysis revealed that an increase in the number of subjects has a significant impact on the industry. Subsequently, a Random Forest model was applied for prediction. After analyzing development trends through visualization, predictive models based on SARIMAX and XGBoost were established. The results indicate that both production and export volumes will show an upward trend. Gaussian Process Regression analysis was conducted on enterprise data from 2019 to 2023. Based on these findings, recommendations are provided with the aim of offering some reference and suggestions for research in related fields.