透過您的圖書館登入
IP:13.59.218.147
  • 學位論文

決策者的風險態度與門檻值對決策結果之影響

The Impacts of the Risk Attitudes and Thresholds of the Decision Makers on Decision Results.

指導教授 : 王敏
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


決策分析可以幫助決策者尋求適當的方案,但由於每種決策工具背後所考量的原理或假設不同,致使在使用不同決策工具評估決策問題時,有時候會導致最佳方案的不一致。然而,決策者對於問題的看法應該一致,不應因為使用的方法不同而有不同的結果。有些決策工具所提供的求解方法很容易就可以判斷出兩個方案之間的優劣關係,但所需之步驟可能較為繁雜,如 PROMETHEE 只需透過決策者給定門檻值以成對的方式比較兩方案之間的優劣,但是當方案數或考慮的屬性數過多時,決策分析所須進行的步驟就會愈來愈繁瑣。而有些決策分析方法可以容易地比較多個方案之間的優劣關係,但在評估決策者的喜好程度可能有些困難,如多屬性效用理論(MAUT) 只需透過決策者的效用函數即可比較出方案之間的優劣,但是評估決策者偏好的效用函數常是很難評估的。 決策者給定門檻值是一件較容易決定的事,但是決策者的效用函數常是不容易評估的;當評估的方案數目很少時利用 PROMETHEE II 進行方案間的評估是方便的,但是當方案數很多的情況下就顯得費時。而且利用 PROMETHEE 來進行決策分析時,不易了解決策者對方案間的優劣程度。所以本研究將利用在使用 PROMETHEE 法時決策者所給定的門檻值與方案評估值之分佈來建立相對應之效用函數,藉由的效用函數來減化 PROMETHEE 面臨多個方案或屬性評估評估過程的繁瑣步驟,而且可以更進一步了解決策者對於方案間偏好的差異程度。

並列摘要


Decision analysis can help decision-makers to find an appropriate alternative. Since there are some differences existed in the principles and assumptions in those decision analysis methods, those difference may cause the inconsistence in the results obtained by those methods. However, a decision-maker should have a consistent viewpoint to the same problem that he/she faced, his-her choice should not be altered by different decision methods. Some decision analysis methods provide easy concept to determine the preference among the alternatives but they required numerous steps to accomplish. For example, by using PROMETHEE, the preference between two alternatives is obtained by comparing the difference between two alternatives and the thresholds provided by the decision makers. Once the number of the alternatives or the attribute increases, the number of steps to compare alternatives pairwise may increase significantly. On the other hand, some decision analysis methods can easily recognize the preference among all alternatives, but the evaluation functions are not easy to obtain. For example, by using multi-attribute utility theory (MAUT), the preference order for all alternatives can be obtained simultaneously, however, the utility functions of the decision makers may be difficult to obtain. It is easy to obtain or assess the thresholds of the decision-makers, but it hard to obtain all the utility functions of the decision maker. When the number of the alternatives is small, it is simple to find the preference order of the alternatives by using PROMETHEE II, however, when the number of the alternatives is large, it will need more time and steps to obtain the preference order. Moreover, only preference order, not the preference degree between alternatives, can be obtained by using PROMETHEE. In this research, we provide a method to approximate the utility functions of the decision makes by the indifference and preference thresholds used in PROMETHEE II. By using the approximate utility function in MAUT all alternatives can not only be compared simultaneously, but also the preference degree among the alternatives can be realized.

參考文獻


6. 陳冠傑,應用近似最佳解績效排序法於多目標設計決策之研究,碩士論文,國立成功大學工業設計學系碩博士班,台南(2004)。
1. Bernoulli, D. (1738). “ Exposition of a new theory on the measurement of risk. ” Econometrica, Vol. 22, pp. 23-36.
3. Behzadian, M., R.B. Kazemzadeh, A. Albadvi and M. Aghdasi (2010). “PROMETHEE: A comprehensive literature review on methodologies and applications.” European Journal of Operational Research, Vol. 200, pp. 198–215.
8. Choi, D.H., S.A. Byeong, and H.K. Soung (2005). “Prioritization of association rules in data mining: Multiple criteria decision approach.” Expert Systems with Applications 29, pp. 867–878.
10. Guitouni, A. and J-M. Martel (1998). “Tentative guidelines to help choosing an appropriate MCDA method.” European Journal of Operational Research, Vol. 1(9), pp. 501-521.

被引用紀錄


張瑜庭(2015)。在風險偏好考量下的改良群體PROMETHEE模型之建構〔碩士論文,淡江大學〕。華藝線上圖書館。https://doi.org/10.6846/TKU.2015.00247

延伸閱讀