在不可逆情況下提升單向選擇決策品質之研究

A STUDY ON IMPROVING THE QUALITY OF ONE-WAY SELECTION DECISION UNDER IRREVERSIBLE CONDITIONS

10.6220/joq.202110_28(5).0003

黃允成(Yun-Cheng Huang)；江旻倩(Min-Qian Jiang)；侯宜璇(Yi-Xuan Hou)；張欣頤(Xin-Yi Chang)

單向選擇 ； 有限母體 ； 無限母體 ； one-way selection ； finite population ； infinite population

品質學報

28卷5期（2021 / 10 / 30）

349 - 360

繁體中文

決策是個人或組織中最重要的選擇行為之一。決策之良窳攸關個人或組織之績效。因此，如何提升決策品質，至關重要。本研究欲探討當各可行方案之良窳分布具隨機性，且選擇過程具不可逆性時，如何進行最佳化選擇之決策問題。本研究建構一最適單向選擇策略模型，探討在有限母體及無限母體兩種情況下，最適參考群體之比例應為多少，方能使找到最適候選方案之期望機率極大化。首先探討有限母體之情境下，最適單向選擇策略之規劃與分析；若可行方案之數量頗多或無法事先估算，則於理論上視為可行方案具相對無限性，因此，以無限母體視之，進而提出無限母體情境下之最適單向選擇策略，以供學術及實務應用之參考。本研究針對此一問題，推導出在不可逆情況下，有限母體與無限母體之最適選擇策略之數學公式，並經由範例分析，驗證該公式之正確性與可行性。經由範例分析得知，當參考比例趨近於0.3679（即1/e）時，找到最適候選方案之期望機率達最大化之目標。

Decision-making is one of the most important choices in an individual or organization. The decision-making is about the performance of individuals or organizations, so how to improve the quality of decision-making is crucial. This study constructs a one-way optimal selection strategy model to explore the optimal ratio of reference group in the case of finite or infinite population, in order to maximize the expected probability of finding the optimal candidate. Firstly, it discusses the planning and analysis of the optimal one-way selection strategy in the context of finite population; if the number of feasible schemes is too large or cannot be estimated in advance, it is theoretically considered as a feasible scheme with relative infinity. Therefore, it is regarded as an infinite population. Furthermore, a one-way optimal selection strategy in the context of an infinite population is proposed for reference in academic and practical applications. To solve this problem, this study derives the mathematical formula of the one-way optimal selection strategy for finite and infinite population under irreversible conditions, and verified the correctness and feasibility of the formula through by numerical example. Through the example analysis, when the reference ratio approaches to 0.3679, the goal of maximizing the expected probability of the optimal candidate is found.

1. Abin, A. A.,Beigy, H.(2014).Active selection of clustering constraints: a sequential approach.Pattern Recognition,47(3),1443-1458.
2. Bearden, J. N.,Murphy, R. O.,Rapoport, A.(2005).A multi-attribute extension of the secretary problem: theory and experiments.Journal of Mathematical Psychology,49(5),410-422.
3. Bruss, F. T.(1984).A unified approach to a class of best choice problems with an unknown number of options.The Annals of Probability,12(3),882-889.
4. Calafiore, G. C.(2017).The dating problem and optimal ordering of sequential opportunities.IFAC-PapersOnLine,50(1),934-939.
5. Chun, Y. H.(1996).Selecting the best choice in the weighted secretary problem.European Journal of Operational Research,92(1),135-147.
6. Jones, B.(2019).Avoiding patterns and making the best choice.Discrete Mathematics,342(6),1529-1545.
7. Wang, Y.,Wang, K.-l.,Liu, Q.-s.(2008).Extension on cut-off rules of sequential observation and selection problem.Systems Engineering-Theory & Practice,28(2),74-81.