Forecasting Model on General Budget Revenue of Regional Finance Based on Dynamic Combination of BP Neural Network

The paper constructs a model which dynamically combines explanatory BP neural network and time series BP neural network to forecast the general budget revenue of regional finance. By using the explanatory BP neural network, the model meets the various factors on the impact of regional financial revenue and by using the time series BP neural network, the revenue´s time correlation is met. Then according to the characteristics of regional financial revenue, adjust the weighting factor. By MATLAB simulation, the forecasting results show smaller deviation than other methods in the references. Its forecasting result is relatively accurate. The model can help solve the forecasting problem of regional financial revenue more effectively.