決策球模式之建構與應用

Abstract

決策者偏好常受到背景資訊的影響，本研究發展一套決策球系統，提供決策者視覺化資訊及相似性分析，將決策資訊視覺化，以輔助決策。 此一決策球系統分為法蘭克運算、對等交換、成對比較、群集分析四模式。 法蘭克運算模式是用於單一方案取捨的決策問題， 對等交換及成對比較模式主要是解決多個替選方案的排序問題，而群集分析模式是應用於替選案的分群問題。 本研究成果可廣泛應用於經營管理決策及財務投資決策等。

Decision makers’ preferences are often influenced by background information. This study develops a Decision Ball system to provide visual context and similarity analysis to help decision makers to reach a better decision. The proposed Decision Ball system includes four types of Decision Ball models: Franklin’s Moral Algebra models, Even Swap models, Pairwise Comparison models, and Classification models. Franklin’s Moral Algebra Decision Ball models solve “Yes” or “No” decision problem. Even Swap and Pairwise Comparison Decision Ball models are for ranking problems with multiple alternatives. Classification Decision Ball models treat group problem. The proposed approach can be applied in a variety of decision problems. For instance, a Decision Ball system can assist decision makers in personal decision-making problem, operational and managerial decision problems, and financial decision problems, etc.