一個模糊多評準決策方法之構建及其應用

Abstract

本文旨在模糊環境下，如何構建一套模糊多評準決策方法。文中提供一套 模糊多評準決策方法的演算法則；該演算法則是依據模糊集合理論與層級 結構分析為基礎所構建而成。模式採兩層級結構分析，第一層用以評估決 策準則之相對權重，第二層為各替代方案在各決策準則下之適合度評價。 並以三角形模糊數與語言變數為兩個主要的觀念，作為準則之「重要性」 與方案滿足各準則之「適合度」評估之用。決策者可利用已設定好的語言 值集或自己主觀認知之模糊數，分別就準則的相對權重，與方案適合準則 之程度進行評分，經適當加總後，得到各方案之綜合評價，稱為模糊適合 指數。文中亦提出一套新的模糊數排序方法，此法最先由Kim與Park教授 ，在一九九○年提出，Kim與Park認為評價者之風險態度值，即樂、悲觀 指數，是由單一決策者在資料輸出階段，依決策者當時的主觀認知，另外 決定該風險承擔值。本文提出另一看法，重新定義決策者風險態度之樂、 悲觀指數，藉由本文之定義，不但可從資料中了解個別決策者之風險承擔 態度值，且可直接由資料本身，直接導出群體決策者之風險承擔態度值； 即群體之樂、悲觀指數，可由原始資料直接決定該值。然後，利用此修正 的方法於模糊適合指數的排序，以供決策之參考。本文亦將所提的模糊多 評準決策之演算法則作進一步的討論與引申。並以生物科技產業為例，說 明如何應用本文所構建之演算法則，評選最適之技術移轉策略。並檢討此 方法在應用上有何優、缺點。 This paper proposes a fuzzy multi-criteria decision making mothod for strategy selection in the area of management under fuzzy environment. An algorithm for strategy selection is proposed. The algorithm is based on the concepts of fuzzy set theory and hierarchical structure analysis. The concepts of hierarchical structure analysis with two distinct levels are used in this paper. The first level is to evaluate fuzzy importance of the decision criteria and the second level is to assign ratings to various strategies under each evaluation criterion. The triangular fuzzy number and the linguistic variable are the two main concepts used in this paper to assess the preference ratings of linguistic variables, 'importance' and 'appropriateness'. The decision makers can either employ an assumed rating set or give his own rating by using the triangular fuzzy number to show the individual perception of the linguistic variables. The reason of using triangular fuzzy number is that it is intuitively easy to be used by the decision makers. We also propose a revised method for ranking fuzzy numbers with index of optimism which was proposed by Kim and Park. Our revised method is based on the stage of data input for computing the total index of optimism in multi-person decision making problem, instead of giving the index of optimism independently by a decision maker on the stage of information output. This new look at the index of optimism reflects the pooled risk- bearing attitude of multi-judge. The index of optimism is determined by the evaluation data conveyed by the decision makers at the beginning of the data input stage, i.e. the risk- bearing attitudes of decision makers are reflected in the data input stage. Finally, we use the revised method to rank the final scores of alternatives for choosing the best technology transfer strategy in the area of biotechnology management.