只要是投資就會有風險，為了降低投資風險而產生了投資組合的概念，投資組合的概念是藉由購買一籃子股票以降低投資風險。如何選擇最佳的投資組合，須考慮投資標的的選取以及資金的分配。當一旦選定了投資組合以後，買賣的時間點也很重要，尤其在股票市場中，股價常常受到不確定性的政治和經濟局勢所影響，股價起伏波動頻繁，難以掌握進場及出場的時間點，因此找出好的買賣時機，降低損失及賺取利潤，在股票投資上，也是非常重要的課題。
本研究使用基因演算法 (Genetic Algorithm)作為挑選股票及資金分配的方法。利用基因演算法每一代交配及突變的演化，找出最佳的股票投資組合；並且利用加權模糊時間序列 (Weighted Fuzzy Time Series)，預測出每一候選的股票投資組合的未來價格，計算出未來可能報酬率，作為基因演算法找出最佳股票投資組合的依據。當選定好一組股票投資組合後，本研究對買賣交易擬定兩種策略：一、定期檢查 二、設定停損、停利點。以每日觀察投資報酬是否達到停損，以及定期比較投資組合報酬率找出更好的投資組合，來判定何時應作出買賣決策，經實驗發現，本方法可有效增加投資報酬率。

Investments in a financial market may incur risk. To reduce the risk in an investment, many portfolio selection methods have been introduced. By buying a set of financial assets, portfolio selection aims at maximizing the return of the investment for a given level of risk. To build an optimal portfolio, one needs to select proper assets and to decide the proportion of the investment for each selected asset. On the other hand, prices of assets vary as time goes by. As a result, the return of the portfolio is also varies with time. Thus, besides the portfolio selection, how to discover the selling and buying points of a portfolio is also a very important issue.
In this study, we develop a genetic algorithm to build optimal stock portfolios. The genetic algorithm uses an evolutionary process to find the portfolio with the highest fitness value as the optimal portfolio. In the proposed method, we use a weighted fuzzy time series to predict the return of the stock portfolio, which is then used as the fitness value of the genetic algorithm. Furthermore, we propose a stock portfolio trading model to find the selling and buying points of the stock portfolio. There are two strategies in the stock portfolio trading model: to periodically check the portfolio returns and to set a stop-loss point for the portfolio. With these two strategies we can find the best selling and buying points for the portfolios. Through experiments on the stocks in Taiwan 50, the proposed method outperforms TAIEX index and Taiwan 50 index in terms of 7-year average return.

1. C.H. Chen, T.L. Chen, H.J. Teoh, C.H. Chiang, “Fuzzy time-series based on adaptive expectation model for TAIEX forecasting,” Expert Systems with Applications, 34 (2008) 1162-1132.
2. D.Z. Cao, S.L. Pang, Y.H. Bai, “ Forecasting exchange rate using support vector machine,” Proceedings of the Fourth International Conference on Machine Learning and Cybernetics, Guangzhou, 18-21 August (2005).
3. E.F. Fama, “Efficient Capital Markets: A Review of Theory and Empirical
Work,” Journal of Finance, Vol. 25, (1970), pp.383-417.
4. Frost , A.J., Prechter Robert R., Jr., “Elliott Wave Principle (10th ed.) ,” Gainesville, GA: New Classics Library, (2005) p. 9.
5. Harry Markowitz, “Portfolio Selection,” Journal of Finance, Vol. 7, No. 1. Mar., (1952), pp. 77-91.
6. Hicks, J. R. and Allen, R. G. D., “A Reconsideration of the Theory of Value,” Economica, 1, (1934)pp. 52-76, 196-219.
7. H.K. Yu, “Weighted fuzzy time series models for TAIEX forecasting,” Physica A, 349 (2005) 609-624.
8. Holland, J., “Adaptation in Natural and Artificial Systems,” University of
Michigan Press, (1975)19880.
9. J.S. Chen, C. L. Chang, J.L. Hou, Y.T. Lin, “ Dynamic proportion portfolio insurance using genetic programming with principal component analysis,” Expert Systems with Applications,35 (2008)273-278.
10. J.S. Chen, J.L. Hou, S.M. Wu, Y.W., C.C., “Constructing investment strategy portfolios by combination genetic algorithms,” Expert Systems with Applications, 36 (2009) 3824-3828.
11. Lamartine Almeida Teixeira, Adriano Lorena Inacio de Oliveira, “ A method for automatic stock trading combining technical analysis and nearest neighbor classification,” Expert Systems with Applications, 37 (2010) 6885–6890.
12. L. A. Zadeh“ Fuzzy sets,” Information and Control ,3(1965) 338–353.
13. Q.H. Wen, Z. Yang, Y.P. Jia, “Automatic stock decision support system based on box and SVM algorithm,” Expert Systems with Applications, 37 (2010) 1015-1022.
14. Robert Rhea, “The Dow Theory,” Fraser Publishing Library,(1932).
15. S.H. Liao, H.L. Ho, H.W. Lin, “Mining stock category association and cluster on Taiwan stock market,” Expert Systems with Applications, 35 (2008) 19–29.
16. S.-M. Chen, “Forecasting enrollments based on fuzzy time series,” Fuzzy Sets Syst.,126, (2002) 367-377.
17. Williams, J.B., “The Theory of Investment Value,” Fraser Publishing Library, Contrary Opinion Library, (1938).
18. 江吉雄(2002)，遺傳演算法於股市選股與擇時策略之研究，碩士論文，中央大學。
19. 李建邦(2010)，在微陣列資料中進行特徵選取與預測基因調控網路之研究，博士論文，台灣科技大學。
20. 林耀堂(2001)，遺傳程式規劃於股市擇時交易策略之應用，碩士論文，中央　　
大學。
21. 吳詩敏(2007)，組合編碼遺傳演算法於投資策略資金分配之應用，碩士論文，中央大學。
22. 吳訂宜(2007)，台股之弱式效率市場檢定，高苑學報第十三卷p191~220，高雄第一科技大學。
23. 周育如(2007)，使用加權模糊時間序列預測匯率，碩士論文，台灣科技大學。

24. 楊孟龍(2000)，類神經網路於股市波段預測及選股之應用，碩士論文，中央大學。
25. 鄭雅仁(1994)，台灣股市弱式效率市場之再驗證，碩士論文，國立台灣大學。
26. 謝劍平，現代投資學-分析與管理，智勝文化，2006。
27. 蘇木春、張孝德，機器學習：類神經網路、模糊系統及基因演算法則，全華科技圖書股份有限公司，2000。