Research On The Utility Of Forecasting GDP By Gray Neural Network Based On Mind Evolutionary Algorithm

Gross domestic product(GDP)is one of the important indicators that reflects the economic development status and overall development level of a country or a region.Its accurate prediction can provide a reliable basis for the government's economic restructuring and economic development.Therefore,research on GDP forecasting has important practical significance.GDP is affected by many factors.The indicator data of these factors presents complex time series and nonlinearity,and the amount of data is relatively small.When dealing with GDP forecasting problems,the accuracy of the the past forecasting methods is not satisfied.The neural network model has a strong learning ability and operats simply.It has good effects on solving the problems of less data and multiple internal collinearity problems in data.So it is Suitable for forecasting GDP.However,in the prediction,there is often a slow convergence rate and it is easy to fall into the problem of local optimal solution.In order to solve the above problems in the neural network model and improve prediction accuracy,the neural network model of mind evolution and optimization is applied to the prediction of GDP in this paper.Based on the data from Hebei Province from 2005to2016,this paper selects the GDP of Hebei province as an explained variable and the ten indicators such as the number of population in Hebei province,fixed investment,general budget revenue of local finance and household consumption and so on as explanatory variables.Based on the grey model,the Back Propagation neural network(BP neural network)is added to the inlaid grey neural network model and series grey neural network models by inserting and connecting in series.Then the mind evolution algorithm is added to the series grey neural network model and the mind evolution method which is used to optimize the gray neural network model is obtained.Gray model,inlaid grey neural network model,series grey neural network model and the mental evolution method optimizes the gray neural network model are used to analyze the predict effect.The results show that the mental evolution method optimizes the gray neural network model model has a better prediction effect in prediction effect.The model can mine the hidden information from a small amount of data,optimize the initial value and threshold of the gray neural network and improve the traditional neural network model with slow convergence and difficulty in obtaining global optimization problems.In the end,the paper further uses the index values of the GDP-related factors in Shanxi Province from 2005 to 2016 to test the effectiveness of the three models.The results show that the mind evolution optimized grey neural network model is used to predict the GDP value of Shanxi Province with the same highest prediction accuracy.The validity of the model is verified.