Guangxi GDP Prediction Model Based on Principal Component Analysis and SSA–SVM

Economic data such as GDP exhibit complex time series and nonlinear characteristics. Predicting these requires consideration of multiple influencing factors, necessitating extensive computation and tedious model training. Traditional methods often struggle to achieve high prediction accuracy in this regard. In light of these challenges, we propose a Support Vector Machine (SVM) model optimized using Principal Component Analysis (PCA) and the Salp Swarm Algorithm (SSA) algorithm. PCA is employed to reduce the dimensionality of GDP influencing factors, eliminating components with low contribution rates and utilizing those with higher contribution rates as input variables for the SVM. The GDP serves as the output variable. Leveraging the optimization capabilities of swarm intelligence algorithms, we further enhance the SVM model’s performance parameters using SSA. Thus, we establish a GDP prediction model based on PCA-SSA-SVM and evaluate its performance on test samples. Results indicate an average prediction accuracy of 95.76%. Compared with the SVM model, PCA-SVM model and PCA-GA-SVM model, the prediction accuracy of the PCA-SSA-SVM model increased by about 15, 9.4 and 4.3 percentage points respectively. These findings underscore the high predictive accuracy of the proposed model, enabling scientifically sound predictions of regional GDP. Finally, to validate the effectiveness of the model, it was applied to GDP prediction for the Beijing area. The results indicate that the model performs well in practice, demonstrating its robustness.