China's export trade structure is undergoing significant changes,with the share of electronic products steadily increasing year by year,and mobile phone products accounting for a large proportion of the electronic product export share.To accurately predict mobile phone export value,a forecasting method based on the Support Vector Regression(SVR)model is proposed.First,monthly data of China's mobile phone export value and its influencing factors since 2012 are selected.Next,data preprocessing is performed to improve the prediction accuracy.Then,Bayesian optimization is applied to select appropriate parameters.Finally,a Gaussian radial basis kernel function is chosen,and the autoregressive integrated moving average(ARIMA)model,SVR model,and Bayesian-optimized SVR model are compared to select the best model for forecasting mobile phone export value.By comparing the actual and predicted values,it is found that the Bayesian-optimized SVR model provides the most accurate prediction of China's mobile phone export value,confirming its effectiveness in forecasting,and offering decision-making support for Chinese mobile phone manufacturers.
维普
万方数据