低利率年代的來臨使得投資理財規劃成為追求最大報酬的主要對策，在眾多投資理財工具中，如何 同時兼顧報酬與風險，以使其取得平衡，顯然為一重要的議題。共同基金是一種受歡迎的投資模式，其 委由專業經理人代為操作，而具有較完善的資訊收集及研判能力、兼顧「較佳報酬」及「較低投資風險」 的雙重特性等優點。另一方面，雲端運算的成本及速度使得上述的投資行為如虎添翼般提供更完善的資 訊、知識等的輔助。 本研究提出一「客製化演進式多目標基金投資組合」的決策模式，亦即從眾多會影響基金的因素中， 以分析層級程序法( analytic hierarchy process, AHP)來客觀量化專家建議及投資者偏好、風險承擔能力、及 其他需求，然後演進式地規劃投資組合，以知名軟體 Lingo 在雲端平台上即時地改善或監測投資行為， 並且協助投資人找到「組合式基金」的滿意建議；期望在可接受的範圍內，能夠達成預期的投資理財目 標。演進雛型法所建置的雛型系統除了提供客製化投資組合的參考決策外，各種管理分析、內涵及投資 原則更可以從系統的不同實證結果及模擬實驗中獲得，以作為未來類似投資組合的參考建議。

The main strategy for facing the coming of low-interest age is finding the financial/ investment plan for maximizing one’s rewards. Using numerous financial/investment tools, how to balance both rewards and risks from the tools in meantime obviously becomes an important issue. Mutual fund is a popular investment tool which entrusts the professional agent to operate one’s capital, and owns several advantages, such as better information collections and judgment ability, dual characteristics for balancing both rewards and risks from fund portfolio, etc. On the other hand, the cloud computing providers which offer services of low cost and high speed can improve the investment behavior for perfect support of information, knowledge, etc. A multi-objective decision model for customizing evolutionary fund portfolio is presented in this study. Numerous factors that affect fund performance are used in the analytic hierarchy process (AHP) for quantifying experts’ suggestions, and investor’s preferences, risk tolerance, and other requirements. Then the model can be solved by a famous software, Lingo on cloud platforms, and used to monitor and improve the investment behavior in real time and help the investor find a satisfied fund portfolio, such that an anticipant goal of finance and investment can be approached within the scope of acceptable risks. Besides the prototype system is developed by the evolutionary prototyping method and can offer the decision-making references of customized fund portfolio, various managerial analyses, implications and guidelines can be obtained from the results of different empirical tests and simulation experiments of the system. Finally, detailed analyses can support decision making under similar situations in future.