| For the purpose of developing an efficient forecasting model for The Corporate Goods Price Index (CGPI), this paper makes many empirical studies of different forecasting models. Through comparing and optimizing these different forecasting models, an efficient model for forecasting CGPI has been selected. This paper has successfully optimized the original forecasting models and significantly improved the forecasting accuracy. Because of the ability to find global optimal parameters and to avoid accidental error, the selected forecasting models not just can be used in forecasting CGPI, but also can accomplish the forecasting job of other fields efficiently.The Corporate Goods Price Index is a comprehensive measure of inflation and economic fluctuations. If the People’s Bank of China hopes to establish and adjust monetary policies promptly and effectively, it is crucial to accurately forecast the CGPI. It is an undoubtedly very challenge task, however, given that a large number of market economic factors could influence the CGPI. To tackle this challenge, this paper aims to provide efficient models to forecast CGPI. Over the past few years, Exponential Smoothing models have performed well in numerous empirical studies especially in the field of economic forecasting. Therefore, Exponential Smoothing models will try to do the forecasting job first in this paper.It is rather difficult, however, to select right parameters during the forecasting process, which often determine the forecasting accuracy, for the Exponential Smoothing models. This study applies three Artificial Intelligence algorithms to optimize parameters, which are generally selected manually, of four Exponential Smoothing models. Finally, the result of an empirical study using the data of Corporate Goods Price Index of China from January2007to September2011show that the optimized Exponential Smoothing models can significantly improve the forecasting accuracy comparing to original Exponential Smoothing models. |
PDF Full Text Request