近年來，由於金融業的發展與進步，讓我們投資的種類越來越多，而且大眾對於投資的觀念也越來越普遍，投資已經成為我們生活的一部分，於是投資理財成為一重要的課題。在眾多的衍生性商品中，股票市場仍然是大眾最青睞的一種投資模式，但是至今的股票市場琳瑯滿目，如何在眾多中挑選出最佳投資組合該是目前最重要的課題，因此本研究希望建立一合理且具邏輯的選股策略，以提供投資大眾使用與建議。本研究以公司基本面、財務面、技術面、市場面及籌碼面的五大構面分別EPS 選股法、本益比選股法、營收選股法、資產報酬率選股法、股東權益報酬率選股法、毛利率選股法、負債比例選股法、RSI 選股法、KD 選股法、MA 選股法、MACD 選股法、股價選股法、漲跌幅度選股法、成交量選股法、主力買賣超選股法、融資融券選股法、外資選股法及投信選股法總共18 個選股法挑選出排名較前的公司利用基因演算法挑出最佳投資組合，其結果發現，基因演算法所選出來的投資組合報酬率都高於大盤，可藉此提供給投資大眾作為選股的準則。

In recent years, due to the development and progress of the financial sector, people have proposed a general idea that both wealth management and investing become part of our daily life, consequently, financial investing is now an issue that cannot be neglected. Among the financial derivatives, the stock market is still favored by most of the people, moreover, the market was filled with an assortment of stocks, as a result, how to select the best portfolio has become a crucial issue, therefor, in this study, we hope to establish a logical and reasonable strategy for stock selection. This study applies genetic algorithm to construct stock investing strategies. We select stock sample by considering five stock-picking criteria including fundamental, financial, technical and market analysis as well as institutional holdings and ownership analysis. We further construct and determine the optimal portfolio, holding period, and market timing. Our findings show larger positive returns in genetic algorithm portfolios than the market. Thus, our results provide more valuable insights on investment strategies.