碩士 / 國立中正大學 / 資訊管理所 / 97 / This study aims to build a neural network model for price forecasting of Japanese eels bred in Taiwan. The back-propagation network was applied to build the forecasting model and another regression model was simultaneously built to compare the prediction accuracy.
Data of the eel breeding industry from 1994 to 2007 was used to train and test the models. The total sample size is 168 months. Twenty-four moving windows were finally used and each moving window consisted of 96 training data and 3 testing data. Three statistics, including Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) were used to evaluate the prediction accuracy. The results indicated that the neural network models outperform the regression models. Implications form the findings are also provided.
| Identifer | oai:union.ndltd.org:TW/097CCU05396019 |
| Date | January 2008 |
| Creators | Yung-Ru Yu, 余泳儒 |
| Contributors | Shin-Yuan Hung, 洪新原 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 79 |
Page generated in 0.0187 seconds