Stock Price Forecasting Using MARS-The Case Study of Formosa Regent Stock
| Autor: | Jheng-Dar Yang, 楊政達 |
|---|---|
| Rok vydání: | 2012 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 100 The research uses a new statistical method –Multivariate Adaptive Regression Spline (MARS) to forecast stock price changes rather than the often used artificial neural network(ANN), which is unable to determine the variable importance and need much more training time. The objective of this study is to investigate the stock price changes taking FORMOSA Regent stock for example. Comparison with past researches using few financial or economic variables, this study sets up MARS model based on forty financial, economic, and other stock variables to forecast stock price changes. It was found that the MARS model forecasted stock price changes, needing time of forecast, and determined the variables importance was better than the back propagation network(BPN) and classification and regression trees (CART).Thus, the forecast results of stock price changes is available by MARS model for investors’ reference. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|