财政收入的组合预测

Combination Forecasting for Financial Revenue

财政收入是政府宏观调控的重要手段,受众多因素的影响,应用组合预测方法来综合考虑多种因素预测财政收入.首先,根据财政收入的组成结构和随时间的变化规律分别建立回归模型与时间序列模型两个线性模型;然后根据财政收入时间序列建立了BP神经网络的非线性模型;最后在此基础上建立组合预测模型,对财政收入预测进行了有益的探索.

Finance revenue is a vital tool for macroscopic regulation and control and is influenced by variety factors.In consideration multifactors effects,the combination forecasting is used to forecast the financial revenue.Two linear forecasting models including multivariate linearity regression and time series are firstly set up before a nonlinear model of BP neural network is established.Based on these three single models, a linearity altering weight combination forecasting model is put forward to study the financial revenue.