Mining the Rules of Investment Using Genetic Programming

• Main Author: 楊蕙憶
• Other Authors: 林文修
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/84499487586496580275

碩士 === 輔仁大學 === 資訊管理學系 === 91 === Nowadays, the investment of stocks is the most popular way to manage finances. Besides the general investors, there are coporations and government in the security market. The proof of the inefficiency of security market and the feature of the asymmetric informations compose of a complicated security market, so it isn’t easy to get abnormal return of investment. Recently years, the value-investing have been puutting in use in the stock market. Many experts of investment, for example Peter Linch, Grahand etc, they use the financial ratios and the value according their experience to construct the rules of investment. They use these rules to find the stocks that have the reasonable price, and they usually could get abnormal return. So, this research will use genetic programming to find rule models by the feature of evolution and the power ability of searching of GP. Combine the general financil ratios, the relative indexes of evaluation model and the suitable value to construct rules. These rules can help investors to find the stocks that have the reasonable price and get abnormal return of investment. This research uses the data of cement industy, food industry and plastics industry of traditional industry to be the research population. The conclusion including: (1) Useing GP to construst the rules really can help investors to find the stocks that have the reasonable price and get abnormal return. (2)Compare the rules combining the general financil ratios, the relative indexes of evaluation model to other models, can get good return of investment. (3)The longer financial report, the information is more adequate and the fluctuation of return of investment is small. In short, this research use the rules evoluted by GP can help effectively the portfolio strategy of investors and get good return of investment. In other words, this research contributes to the future research on GP and financial field and also has application in practice.