Dynamic Evaluation and Prediction for Macroeconomic Based on IGA-LSSVM

Among all statistics, government statistics plays an extremely important role in socioeconomic activities, while the continuous changes of external factors bring new challenges to and have a great impact on quality statistics. With rapid development economy maturity market system, demand for from walks life has become more vigorous. In order effectively improve accuracy macroeconomic outflow forecasting further secure economic production, weighted least squares support vector machine (LS-SVM) optimized with immune genetic algorithm (IGA) is proposed build model. For nonlinearity, time-varying, complexity strategy function LS-SVM, then IGA introduced optimize improved LS-SVM kernel parameter δ regularization γ. Finally, experimental analysis conducted using emergence data history. The results show that maximum relative error model 2.763%, minimum 0.705%, average 1.3298%, faster convergence, stronger generalization ability, higher than other models.