Study On Algorithm Of Statistical Forecasting Model And Its New Development

This thesis makes a research on the algorithm of Statistical Forecasting Models and some of its new developments today. It introduces not only traditional forecasting models,but also some popular models which represent the current trend in this field.. The thesis improves some of models. It introduces both traditional forecasting models and some popular models which represent the current trend in this field; meanwhile it proposes some improvement plans. All the models in this thesis are programmed by MATLAB due to its effectiveness of forecasting, and are applied to the case studies of Chinese economy.This thesis contains five chapters:Chapter one introduces the concept and the present situation of statistical forecasting, as well as the functions of MATLAB.Chapter two analyses the traditional forecasting models, Regress Frecasting Model and Time Series, Grey Model and Neural Networks Model. These models are applied to the study of China's economy with the use of MATLAB.Chapter three analyses seasonal adjustment models: X-11 and X-12-ARIMA. It brings forward an update plan aimed at improving X-12-ARIMA's disability to take the Spring Festival factor into account adjust to the Spring Festival factor in China.Chapter four introduces combining models and the new method-AFTER, which is applied to Chinese economic forecasting. Chapter five makes a conclusion and provides some relevant advices.