Tax revenue forecast for the existence of nonlinearity,instability and economic factors that affect multiple complexities,the author proposed to use the method of least squares support vector regression machine to predict the tax revenue of Guangdong conghua,and establish the mathematical model.Support vector machine is the core of statistical machine learning,which is based on VC dimension theory and structural risk minimization theory,and has been widely used in the fields of pattern recognition,function fitting and regression prediction.In its application,the choice of kernel function and the setting of parameters have great influence on the model of SVM.Therefore,this paper adopts the intelligent algorithm to select the parameters automatically in the support vector machine.The commonly used intelligent optimization algorithm has Particle swarm optimization algorithm,genetic algorithm,grid search algorithm and so on.The main work of this paper is as follows:Firstly,as the parameters of the regularization parameter C and radial basis kernel function g in the support vector machine directly affect the prediction effect of the algorithm model,this paper uses the particle swarm optimization algorithm,the genetic algorithm and the grid search algorithm for the regularization parameters C and the radial parameters g of the kernel function are optimized.Secondly,the principal component analysis is carried out on the indicators that influence the tax revenue,eliminate the redundant variables between the indicators,reduce the dimension processing,and use the processed data as the input data.Finally,the optimal algorithm C is 1.1859 and the best g is 0.1 by the particle swarm optimization algorithm.The best obtained by genetic algorithm C is 4.4529 and the best g is 0.018172.The best result is obtained by the grid search algorithm C For the 16,the best g for the 0.0039063.The average absolute percentage error and mean square percentage error of the support vector machine model with particle swarm optimization are the smallest,and the tax forecasting model achieves a good effect.Followed by the genetic algorithm to optimize the support vector machine model,and finally the grid search algorithm to optimize the support vector machine model.At the same time,this paper also compares the support vector machine models with no principal component analysis and the support vector machine models for principal component analysis.The results of the principal component analysis of the tax data are better and the accuracy is higher,and enrich the tax forecasting research methods. |