碩士 / 國立屏東教育大學 / 應用數學系 / 98 / The study object of the research is SSE Composite Index. Regression analysis and time series are used to forecast the closing price on the next day and compare the forested results of all models. The search starts from March 2009 to March 2010; there are 263 pieces of data in total. The technical indicator calculated from the 241 pieces of data from March 2009 to February 2010 set up multiple regression model to forecast the closing price on the next day; daily closing price of the same duration are used as time series data to set up ARIMA model. The 22 pieces of data in March 2010 are used as test samples to evaluate the forecasted results and the indicators for evaluating forecasting ability are RMSE, MAE, and MAPE as error differentiating indicators.
The results show that in the two models set up by regression analysis and time series, the three error differentiating indicators of ARIMA model are all smaller than those of multiple regression model, so the forested results of ARIMA model are better than those of multiple regression model.
| Identifer | oai:union.ndltd.org:TW/098NPTT5507006 |
| Date | January 1900 |
| Creators | Tai-An Chen, 陳泰安 |
| Contributors | Kuo - Kang Chang, 張國綱 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 58 |
Page generated in 0.0015 seconds