| Summary: | 碩士 === 國立臺灣科技大學 === 資訊管理系 === 92 === The idea today which we manage finances in the investment popularizes day after day, so each kind of investment tool was connected with the most of modern people’s life. In the future, professional management in investment and finance will become a trend, due to mature in stock markets, boom of the listed companies and complex correlation in global financial markets. The mutual fund has the superiority , since there are professional management, and contains many investments to enable it with the advantage of sharing risks and enjoying profit stably. In the past, the question which the investor faced is too many stocks, however the present is too many fund. In recent years, the combined type fund started to develop slowly in domestic, which take the fund as the investment sign and helps the investor by the professional angle to solve the problem to choose fund.
On the other hand, along with computer information and artificial wisdom developed, many questions unable to be solved by traditional method all have the unprecedented development. This research attempts to utilize genetic algorithm(GA) and ant colony system(ACS) to establish decision-making of fund assignment for combined type fund, and develops a policy-making pattern refered by the investor to construct the combined type fund.
The empirical results are as follows:
1. In respect of the convergence of Sharpe ratio, the convergent results from ACS and GA are almost identical.
2. The reward rates respectively from the decision-making constructed by ACS and GA differ not much. But they are both distinctly better then fund equal distribution and holding weighting index.
|
|---|
