Application Of Artificial Neural Network Algorithm In GDP And CPI

International economic competition is becoming more and more fierce.In order to ensure our country standing in an invincible position in international competition,so we must ensure the operation of the macro economy healthy and stable.GDP growth rate reflects the overall level of the country's economic situation,and CPI index directly affects the purchasing power of residents.In the formulation of macroeconomic policies,we must study the historical data,and find the relation between them from the historical data,so as to provide guidance for policy making.The historical data of GDP growth rate and CPI index have complex time series and nonlinear characteristics.The artificial neural network algorithm has good nonlinear fitting ability,and has been used widely in dealing with nonlinear problems.As the artificial neural network algorithm which used widely,both BP neural network algorithm and SVM algorithm have good nonlinear fitting ability,however they have their own shortcomings.Aiming at the deficiency,this paper has done the following work.1.BP neural network algorithm is easy to fall into local minimum,so the PSO-BP model algorithm is proposed.In this method,the PSO algorithm is used to optimize the weights and thresholds of BP algorithm,so as to avoid the local minimum of BP algorithm.2.Parameter selection of SVM algorithm affects the performance of the model directly,so how to choose the appropriate parameters are crucial.In this paper,for the SVM parameters are optimized with the global GWO algorithm using gray good searching ability to improve the prediction accuracy of SVM models.3.The principal component of PCA can reduce the dimension,and retain the most information of the original data.In this paper,the PCA algorithm is used to reduce the experimental data in order to improve the prediction accuracy of the model.The GDP growth rate and CPI index were fitted and predicted by PCA-PSO-BP model.The experimental results show that the PCA-PSO-BP model has higher fitting accuracy and smaller mean square error than PCA-BP model and PSO-BP model.In addition,the GDP growth rate and CPI index of the PCA-GWO-SVM model are also predicted.The results show that the PCA-GWO-SVM model has higher fitting accuracy and smaller mean square error than PCA-SVM model and GWO-SVM model.Compared with the experimental results of PCA-PSO-BP model and PCA-GWO-SVM model,the PCA-GWO-SVM model has a smaller mean square error.