本研究主要目的是應用遺傳演算法(Genetic Algorithms, GA)協助投資人在瞬息萬變的股票市場，建構一個捕捉股價趨勢轉折點(turning point)的最適技術指標(Technical indicator)組合，以及資金分配比率的模型，藉以提升獲利與降低風險。本研究以GA的演化尋優特性，以二階段方式發展台灣股價趨勢轉折點與資金配置的波動捕捉模型；第一階段是利用股價具有波段漲跌的特性，以GA挖掘股價轉折點，藉以確認每個技術指標組合的最適值；最後，採用技術指標值與其平均線的「相對」數值；第二階段為發掘每個技術指標不同的買賣「權重」(Weights)，藉由權重做為資金分配比率，做為股票的交易策略。本研究實驗結果發現：(1) 本研究的GA模型具有追蹤股價波動（volatility）與趨勢轉折點(turning point)的能力。實驗顯示GA模型在現貨交易與融資融券交易策略，都比買進持有（buy and hold）策略的報酬率高，而且融資融券又遠比現貨交易策略的獲利性大。(2) 在空頭市場(Bear Market)，本研究模型明顯比買進持有策略在虧損實驗期的虧損較少，但在獲利實驗期時卻又獲利較大，顯示本研究模組具風險控制的能力。(3)本研究創新採用GA進行股價趨勢轉折點的技術指標組合的最佳化，以及技術指標的「相對」數值與「權重」資金配置的萃取，確實能在具有複雜與混沌特性的股票市場中，降低其不確定性與隨機性。

This research applied the genetic algorithms (GA) to construct a recommendation model, and this model assist investors to make investment decision. This model utilize stock price have wave band characteristic of ups and downs; find turning points of the stock price. And then cooperate with fund allocation to use, judge whether stocks should be bought or sold. The result of this research can be summed up for the following several points: (1) the profit of margin trading is higher. (2) In comparative analysis, the profit of this research model is higher than to buy and hold. (3) In the bear market, this research model is unlikely to get the positive profit, but it will be less than the ones that bought holding but suffer the loss. (4) This model apply “Relative Value” and “Weight”. This model can get the higher positive profit and control risk.