本研究之主要目的,係建立可模化人類偏好結構與多目標決策行為之理論
架構與方法論.與以往研究不同處,在於本研究考量現實問題之需要,同時
提出重要程度求解之靜態,動態分析.其次,本研究克服一般模糊測度於實
務應用時所遭遇之困難,利用實驗設計技術發展資料收集程序,並以基因演
算法進行求解程序設計.再者,以個人行為為基礎來構建動態權重與重要程
度之求解方法,而且,亦將資訊提供之效果納入考慮.

The main objective of this dissertation is to develop a
theoreticalframework and methodology that can model the human
preference structureand capture multiple criteria decision-
making. The work presented inthis dissertation is different from
previous studies in a number ofsignificant aspects. First, the
static and dynamic analyses regardingthe importance of
evaluation criteria were conducted to denote decision-makers''
preference structure. Second, this study surmounted the
implementationdifficulties of general fuzzy measures in practice
by experimental designtechnology. In addition, the
identification procedure for general fuzzymeasures was proposed
by applying genetic algorithms. Third, the individualbehavioral-
based mechanism was established to assess the dynamic weightsand
grades of importance for influential criteria. Furthermore, the
effectsof external information were investigated as well.