透過您的圖書館登入
IP:18.219.63.90
  • 學位論文

類神經網路應用在台灣股價指數期貨預測之研究

The Application of Artificial Neural Network for Studying the Taiwan Stock Index Future

指導教授 : 吳肇銘
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


民國87年7月21日台灣期貨交易所正式推出台灣發行量加權股價指數期貨契約(簡稱台股指數期貨),台股指數期貨上市時間最久,擁有的成交量最多,因此,本研究以台股指數期貨做為主要的研究對象,嘗試找尋出最適宜及準確的預測模型,來創造期貨投資者最大的獲利空間。   本研究選取台指0512~0604五個期指商品之前一月份商品結算後至該期指商品最後結算日前,選取各項研究變數之15分鐘資料(最低價、最高價、開盤價、收盤價、成交口數、買價成交筆數、賣價成交筆數、委買單未成交筆數、委賣單未成交筆數、委買單未成交口數和委賣單未成交口數),作為實驗數據,並設計三個預測模型,將相關的變數輸入類神經網路系統中,完成以類神經網路建構台指期貨(最高價/最低價)的預測模型。   實驗測試階段預測準確度為預測模型三>預測模型二>預測模型一,但是進行三次的盤中實測,其實測的結果與測試結果有些差異,既使如此,本研究之預測模型對台股指數期貨預測,仍有不錯的預測能力。

並列摘要


Taiwan Stock Exchange Capitalization Weighted Stock Index Futures (TAIEX) launched by Taiwan Futures Exchange (TAIFEX) in July 21st, 1998, which was the one to enter the market for the longest time, and to have the biggest turnover volume as well. Thus, TAIEX is the major object of this study. The purpose is to make a profit space for future investors by trying to find out the appropriate and accurate module. The five futures transactions of TAIEX from 0512 to 0604 were selected at this study, and the data of every number of variables for 15 minutes were selected as well for experiment statistics. Furthermore, the three divinable modules were designed to include the related number of variables into analogical affiliated network system for TAIEX forecast. The sequence of the experimental accuracy is as followings, the module 3>the module 2>the module 1 . Even though, there are some differences of experiments between realistic and forecast results, but the forecast ability of this study module is still good for TAIEX.

參考文獻


8. 李惠妍(2003),「類神經網路與回規模式在台股指數期貨預測之研究」,國立成功大學高階管理碩士在職專班碩士論文。
1. Bergerson,K. and Wunsch,D.C.,(1992), “A Commodity Trading Model Based on a Neural Network-expert System Hybrid”,IJCNN-91,I,pp.289-293.
3. Fama, E. F.,(1965),“The Behavior of Stock Market Prices” Journal of Business, Vol.33, January, pp.34-105.
6. Fama, E. F.,(1970),“Efficient Capital Market: A Review of Theory and Empirical Work,” Journal of Finance, Vol.39, January, pp.226-241.
7. Kendall, M.,(1953),“The Analysis of Economic Time Series, PartⅠ: Prices,” Journal of the Royal Statistical Society, Vol.96, pp.11-25.

被引用紀錄


吳政修(2008)。應用馬氏田口系統於股價預測之研究-以台灣電子類股為例〔碩士論文,中原大學〕。華藝線上圖書館。https://doi.org/10.6840/cycu200800284

延伸閱讀