Artificial Neural Network-based Economic Growth Forecast

Based on existing studies of economic forecasting methods,an Artificial Neural Network(ANN)-based algorithm is developed to make economic forecasting more realistic in this thesis.Through selection of the appropriate ANN algorithms an economic forecasting model is established .First,this thesis summarizes and studies characteristics,steps and error-checked in the economic forecasting.It is revealed that the traditional methods are becoming unfit for the modern economic management need,because of economic management be more complicated nowdows.Ann is a system with the function of vast scale and parallel information disposal. It can dispose nonlinear function and relation more easily and efficiently.so the ANN-based economic forecasting has been found encouraging.Secondly, this thesis looked back on the history of ANN development,and introduced structure and character of ANN. Discussing on using ANN in economic management,this paper points out that ANN is not only used in area of engineering project,but also in area of economic management.Thirdly,this thesis looked back on the development of increment theory,and this theory is about classical economic growth theory and neoclassical economic growth theory and endogenous growth theory.Finally,this thesis used increase of three types of production as input,and used GDP as output. Then the paper selects the good algorithm and built the model of ANN.Using statistics data,we carry out quantitative analysis The result from the case study shows that ANN-based economic forecasting method is useful and effective and can be a reference for dicision-makers of economic.