碩士 / 輔仁大學 / 金融與國際企業學系金融碩士班 / 106 / This paper use China's Shanghai A share index as the research target. The research variables include China's manufacturing purchasing managers' index, Monetary Aggregates M1, Monetary Aggregates M2, the exchange rate of RMB against the U.S. dollar , one-year interest rate, foreign exchange reserves, and consumer price index,etc. The research period is the monthly data for 11 years from January 2007 to December 2017. And then use these research variable data in Long Short-Term Memory Model to construct the Shanghai A share index prediction model. Last, the seven Macroeconomic Variables from January 2007 to December 2016 will be used to forecast the trend of the monthly Shanghai A share index from January 2017 to December 2017.
The prediction model constructed in the first method is used the cross-analysis method, and the predicted result hit rate is 72.73%, and the average MSE in the test period is 0.0010.The prediction model constructed in the second method has a trend hit rate of 63.64%, and the average MSE in the tested is 0.014, indicating that the LSTM prediction model constructed with the overall economic indicators has good predictive power.
| Identifer | oai:union.ndltd.org:TW/106FJU00214002 |
| Date | January 2018 |
| Creators | CHENG,YU-CHIEN, 鄭余建 |
| Contributors | HAN,CHIEN-SHAN, 韓千山 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 41 |
Page generated in 0.0011 seconds