| Summary: | 碩士 === 輔仁大學 === 資訊管理學系碩士在職專班 === 104 === The mainly goal of the research is searching for the best top and bottom of box applying Genetic Algorithms （GA）, developing an useful stock exchanging strategy system, along with Gene Expression Programming （GEP）, to evaluate investment profit.
Therefore, the research is devided into two parts. First of all, using GA to search the best top and bottom of box from historical data automatically, after that, using these perfect ratio box type to invest future stock market, observing its effectiveness of investment. Secondly, using GEP analysing ability to develop strategy of exchange, adding the perfect ratio box data as stop loss system, examining and analysing its effectiveness of investment. I hope the outcomes of the research could create an exchanging system balanced profit and risk, providing investors to use practically, and contributed in exchanging strategy of research.
In experimental conclusion, we found that the box theory isn't appropriate to those low-movement stocks. Besides, there's no perfect box model in stock market, considering perfect flexing box would go with the trend. When the stock market rose drop large, the optimum box becomes large, otherwise small. I found the GEP would gets profit steadily while the stock market goes up dramatically, and reduces loss while it goes down. The effectiveness of invest risk is GEP in the four model we examined. GEP could control risk effectively and safely, obviously.
|
|---|
