隨著資訊的便利,網路的發達,投資人資訊取得更為快速,過去常經由報紙、新聞、投顧,甚至是小道消息進行操作的時代已經過去了,相較於過去,現在投資人擁有更多的資訊,能夠理解數據、判斷數據並且利用數據已經是新的趨勢,而投資人都期望著能夠預測未來走勢,但市場變化多端,任何一個事件都將產生變化,因此投資人大多都無功而返,而投資人了解了指數預測的困難度,而進而試圖利用市場消息在市場上低買高賣,但在網路普及的現今,消息都輾轉他人之手,無法透過消息獲利,而投資人需要利用龐大的數據進行判斷,但如何使龐大的數據幫助投資人做判斷將是一大難題。 模糊決策樹(Fuzzy Decision Tree)是結合模糊理論和決策樹的新演算法,而在近年陸續被廣泛使用,以少數資料便能產生規則進行判斷,對於股市的瞬息萬變,模糊決策樹能只需要少許資料變能做判斷,透過不斷的修正,進而達成預測準確的目的,本研究使用模糊決策樹來預測未來指數走勢發現預測點數方面誤差甚大,預測漲跌情形也不甚理想只有57.4%的準確率,但經過筆者改變原本變數,並設計為新的模型,模型預測準確率高達93.44%,其中的判斷規則由過去走勢所產生,所形成的規則庫幫助投資人做出投資決策,進而能適時且靈活的改變交易策略。
In the past, investors with limited information to make investment decisions, but with Internet facilities and information made easily, the original investment decision is no longer available. So investors must use a lot of information to make investment decisions. However, with all the news on the market affect investment decisions, investors tend to be influenced by messages and make transactions, often resulting in a loss, investors therefore disappear from the market. Today, the use of large amounts of data analysis has become a trend, but in the course of the analysis still has many problems. Fuzzy decision tree is a combination of fuzzy theory and decision tree. It has been widely used in recent years. Fuzzy decision tree can be generated from a small amount of data in the rules. Apply to the rapid changes in the stock market can correct the error quickly, and reached the purpose of accurate prediction. In this study, the fuzzy decision tree to predict the future movement of the Index found that the prediction error is very large number of points. Change the situation forecast only 57.4%. In the present, in this study, the author changed the original variables, and designed for the new model. Prediction accuracy of the model is 93.44%. The trends are from the previous judgments rule, we expected to help investors make correct investment decisions and able to make right trading strategies immediately.