Forecasting is an instrument to infer the trace and direction of the development of a business, and it receives much concern in our daily lives. Recently, more and more forecasting models have been set up. But there still exists a problem: the forecasting based on a very small sample. In this thesis, the author puts forward a semi-parametric regression model for forecasting, based on data simulation and non-parametric theory. Non-parametric estimation has some application on forecasting in foreign research, but domestic researchers seldom apply it in this area. This paper focuses on the estimation of semiparamatric regression model, and puts forward the forecasting steps. At last, applies it into forecasting...
|