The Functional-Coefficient Autoregression Forec AST Model Of Per Capita GDP Of China

The GDP is an important guide line which measures national economic power and economic scale. It is also an importance basis to establish macroscopical control policy. The per captial GDP is the most important comprehensive guide line which presents an economic power of nation, develop level and living standard. It not only considered the size of economic total amount, but also combined the quantity of population, which widely used for an evaluation and compares a nation's economic develop level. We get the result about the model to the per captial GDP of China forecast by using the linear autoregressive model and the function-coefficient autoregressive model.1. The linear autoregressive model is applied to the per captial GDP of China, and the parameter of the model is estimated by the least square estimation , the best way is used to make forecasting.2. Based on the non-parametric functional-coefficient autoregressive model and non-parametric estimation theory, polynomial spline is adopted to establish the functional-coefficient autoregressive model.3. The fitting value and forecasting value of the liner autoregressive model and the nonparametric function-coefficients autoregressive model are compared, the result shows that the nonparametric function-coefficients autoregressive model can give better results than the linear autoregressive model for the prediction of the per captial GDP of China.