基于组合预测模型的中国对俄进出口额预测

摘要

In this paper, in order to more accurately predict the development trend of the bilateral trade between China and Russia, we first analyzed systematically the domestic and international situation for the development of the Sino-Russia trade, used the multiple regression model and neural network model to respectively build the linear and nonlinear forecasting model of the export and import value between China and Russia, then established the combination forecasting model for the trade volume between the two using the secondary optimization algorithm and at the end, verified that the combination forecasting model was an effective method of optimization and filtering in the forecasting of the trade value between China and Russia.%为了准确预测中俄双边贸易发展趋势,首先系统剖析了中俄贸易发展面临的国际国内形势,利用多元回归模型和神经网络模型对中国对俄出口额和进口额分别建立了线性和非线性预测模型,进而采用二次优化算法构建了中俄贸易额的组合预测模型,预测结果表明组合预测模型为中俄贸易额预测提供一种有效的优化和筛选方法,为科学规划中国对俄口岸和深化中俄全面战略协作伙伴关系提供依据.