第五章 結論
5.3 未來方向
本研究尚有許多可以發展的空間，目前的交易決策系統是初步的架構，將有 許多的部份都有待調整，後續發展如下說明:
1. 在本研究的買賣點決策能夠再加上成交量的資料作為研判，因為在股市當中 量是價的先行指標。一般來說，量大價漲、量縮價跌，而量價背離的時候，
將可能是股市反轉訊號。
2. 本交易決策系統採用十種技術指標，往後的研究可以加上更多不同的指標來 研究，以及考慮到融資、融券和三大法人的持股指標。
3. 本研究建構的動態移動平均線雖然已經改善了靜態移動平均線的缺失，但是
尚未考慮成交量的差異，後續研究可以與量能移動平均線的結合。
4. 在股票和資金比例的平衡系統，可以再加上 AI、TAPI、OBOS 等指標，作 為大盤目前動向的研判，將有助於資金的配置。
5. 調整交易系統的歸數函數，以達到推論的結果最佳。
6. 除了考慮單一個股的操作，可以加上投資組合理論來操作，以達到降低風 險。
附錄 技術指標公式
在本研究當中，總共使用了十種技術指標，包括 RSI, BIAS, K, D, WS, QRSI, QBIAS, QK, QD, QWS，其中 RSI, BIAS, K, D, WS 的公式如下，而 QRSI, QBIAS, QK, QD, QWS 的公式將使用成交量替代股票價格輸入所得到的指標。
l RSI:
RSIn RS
− +
= 1
100 100
其中， Down RS = Up
∑
== n
i n
RiseSum Up
1
∑
== n
i n
FallSum Down
1
日內上漲點數的總和 RiseSum =在n
日內下跌點數的總和 FallSum=在n
l BIAS:
n ave n
n P
P BIAS = P −
其中，
∑
=
= n
i i
ave n
P P
1
l K, D:
3 3
*
2 K 1 RSV Kn = n− +
3 3
*
2 n 1 n
n
K
D = D − +
其中， Highest ice Lowest ice
ice Lowest ice
Close
RSV Pr Pr
Pr Pr
−
= −
日內的最高價 ice 在n
Highest Pr =
日內的最低價 ice 在n
LowestPr =
日內的收盤價 ice 第n
ClosePr =
l WS
Highest ice Lowest ice ice Close ice
Higheat WSn
Pr Pr
Pr Pr
−
= −
其中，Highest Price =在n日內的最高價 日內的最低價 ice 在n
LowestPr =
日內的收盤價 ice 第n
ClosePr =
參考文獻
[1] Andrew Colin, Citibank, London, Solving Ration Optimization Problems with a Genetic Algorithm, 1993 Advantage Technology A Monthly How-To Newsletter for Develops, Volume 2.
[2] Baba N., Inoue N., Asakawa H., Utilization of neural networks and GAs for constructing reliable decision support systems to deal stocks, IJCNN 2000 Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks, Volume 5, Page (s): 111 – 116.
[3] Baba N., Asakawa H., Sato K., Application of techniques of computational intelligence for constructing reliable decision support systems, IJCNN 1999 International Joint Conference on Neural Networks, Volume 6, Page (s): 3856 – 3859.
[4] Baba N., Asakawa H., Inoue N., Application of soft computing techniques fo r dealing Tokyo stock exchange prices indexes (TOPIX), 1999 Third International Conference on Knowledge-Based Intelligent Information Engineering Systems, Page (s): 542 – 545.
[5] Baba N., Handa H., Utilization of neural network for constructing a user friendly decision support system to deal stocks, 1995 Proceedings IEEE International Conference on Neural Networks, Volume 2, Page (s): 818 – 823.
[6] Casimir C., Klimasauskas, Developing a Multiple MACD Market Timing System, 1993 Advantage Technology for Developers, October – November – December Special Issue, Volume 2.
[7] Fu Kai, Xu Wenhua, Training neural network with genetic algorithms for forecasting the stock price index, 1997 ICIPS IEEE International Conference on Intelligent Processing Systems, Volume 1, Page (s): 401 – 403.
[8] Jin Li, Tsang, E.P.K., Investment decision- making using FGP: a case study, 1999 CEC Proceedings of the Congress on Evolutionary Computation, Volume 2.
[9] Jian Chen, Dong Xu, An economic forecasting system based on recurrent neural networks, 1998 IEEE International Conference on Systems, Man, and Cybernetics, Volume 2, Page (s): 1762 – 1767.
[10] Jung-Hua Wang, Shiuan-Ming Chen, Evolutionary stock trading decision support system using sliding window, Evolutionary Computation 1998 Proceedings IEEE World Congress on Computational Intelligence, Page (s): 253 – 258.
[11] Jung-Hua Wang, Shiuan-Ming Chen, Jia- Yann Leu, Stock trading decision support system using a rule selector based on sliding window, 1997 IEEE International Conference on Systems, Man, and Cybernetics, Computational Cybernetics and Simulation, Volume 1, Page (s): 559 –564.
[12] Jung-Hua Wang, Jia-Yann Leu, Dynamic trading decision support system using rule selector based on genetic algorithms, 1996 Neural Networks for Signal Processing VI, Proceedings of the IEEE Signal Processing Society Workshop, Page (s): 119 – 128.
[13] Jung-Hua Wang, Jia-Yann Leu, Stock market trend prediction using ARIMA-based neural networks, 1996 IEEE International Conference on Neural Networks, Volume 4, Page (s): 2160 – 2165.
[14] Kimoto, T., Asakawa, K., Yoda, M., Takeoka, M, Stock market prediction system with modular neural networks, 1990 IJCNN International Joint Conference on Neural Networks, Page (s): 1 – 6.
[15] Kuo, R.J., Lee, L.C., Lee C.F., Integration of artificial neural networks and fuzzy Delphi for stock market forecasting, 1996 IEEE International Conference on Systems, Man and Cybernetics, Volume 2, Page (s): 1073 – 1078.
[16] Lim M.- H., Wuncsh D., A fuzzy perspective towards technical analysis-case study of trend prediction using mo ving averages, 1999 Proceedings of the IEEE / IAFE Conference on Computational Intelligence for Financial Engineering, (CIFEr), Page (s): 179 – 182.
[17] Pantazopoulos K.N., Tsoukalas L.H., Bourbakis N.G., Brun M.J., Houstis E.N., Financial prediction and trading strategies using neurofuzzy approaches, 1998 IEEE Transactions on Systems, Man and Cybernetics, Part B, Volume 28, Issue:
4, Aug., Page (s): 520 – 531.
[18] Setnes M., van Drempt O.J.H., Fuzzy modeling in stock- market analysis, 1999 Proceedings of the IEEE/IAFE Conference on Computational Intelligence for Financial Engineering (CIFEr), Page (s): 250 – 258.
[19] Shoaf J., Foster J.A., The efficient set GA for stock portfolios, 1998 IEEE World Congress on Computational Intelligence, The IEEE International Conference on Evolutionary Computation Proceedings, Page (s): 354 – 359.
[20] Talluru L.R., A fuzzy expert system for mutual fund selection, 1997 IEEE International Conference on Systems, Man, and Cybernetics, Computational Cybernetics and Simulation Volume 1, Page (s): 680 – 685.
[21] Tony and Guido Deboeck, Genetic Optimization of Neural Nets for Trading, 1992 Advantage Technology for Developers, A Monthly How-To Newsletter for develops, Volume 1.
[22] Wen-Shiu Lin, Jiah-Shing Chen, Ping-Chen Lin, A study on investment decision- making model: genetic algorithms approach, 1999 IEEE SMC Conference Proceedings. IEEE International Conference on Systems, Man, and Cybernetics, Volume 1, Page (s): 1049 – 1054.
[23] Yang Yiwen, Liu Guizhong, Zhang Zongping, Stock market trend prediction based on neural networks, multiresolution analysis and dynamical reconstruction, 2000 Proceedings of the IEEE/IAFE/INFORMS Conference on, Computational Intelligence for Financial Engineering (CIFEr), Page (s): 155 – 156.
[24] 王鍾億，類神經網路投資決策支援系統投資策略之研究，成功大學企業管 理研究所碩士班論文，民國八十七年。
[25] 王隆盛，運用啟發式演算法於股價指數套利投資組合模型之建立，交通大學 資訊管理碩士班論文，民國八十五年。
[26] 卞志祥，台灣加權股價指數投資組合之基因演算法建構模型，交通大學資訊 管理研究所碩士班論文，民國八十五年。
[27] 吳振坤，類神經網路在台灣股價行為上的應用，交通大學資訊管理研究所碩 士班論文，民國八十年。
[28] 岑英勤，智慧型決策系統運用於台灣股票市場技術面分析之研究，交通大 學資訊管理研究所碩士班論文，民國八十二年。
[29] 林萍珍，遺傳演算法在使用者導向的投資組合選擇之應用，中央大學資訊 管理研究所碩士論文，民國八十七年。
[30] 紀桂銓，類神經網路對股市反轉點之學習與預測應用，交通大學資訊管理 研究所碩士班論文，民國八十二年。
[31] 陳照憲，基因演算法技術交易法則獲利績效-台灣股票市場實證研究，雲林 科技大學企業管理研究所碩士班論文，民國八十二年。
[32] 揚孟龍，類神經網路於股價波段預測及選股，中央大學資訊管理研究所碩 士論文，民國八十九年。
[33] 曾恩博，類神經網路於股價預測與資金配置之應用，中央大學資訊管理研 究所碩士論文，民國八十八年。
[34] 曾淑青，運用類神經網路於台灣股票市場價量關係的預測與分析，交通大 學資訊管理研究所碩士班論文，民國八十三年。
[35] 張振魁，以類神經網路提高股票單日交易策略之研究，中央大學資訊管理研