A novel method to use fuzzy soft sets in decision making based on ambiguity measure and Dempster-Shafer theory of evidence: An application in medical diagnosis
A novel method to use fuzzy soft sets in decision making based on ambiguity measure and Dempster-Shafer theory of evidence: An application in medical diagnosis
复制标题
基于模糊性度量和 Dempster-Shafer 证据理论的模糊软集决策新方法：在医学诊断中的应用
DOI：
10.1016/j.artmed.2016.04.004
复制
发表时间：
2016-05-01
影响因子：
7.5
通讯作者：
Deng, Yong
中科院分区：
文献类型：
--
作者：
Wang, Jianwei;Hu, Yong;Deng, Yong
目的：最近，基于模糊软集的决策引起了越来越多的兴趣。尽管已经做了很多工作，但它们无法提供结果的不确定性或确定性。管理不确定性是决策中最重要和最艰巨的任务之一，尤其是在医学领域。在本研究中，我们改进了基于模糊软集的决策中减少不确定性和提高选择决策水平的性能。方法和材料：我们利用两种适当的工具（模糊性度量和Dempster-Shafer证据理论）来改进基于模糊软集的决策。我们提出的方法由三个过程组成：首先，通过模糊度度量获得每个参数的不确定度；接下来，根据第一步得到的各参数的不确定度，构造针对各参数（或证据）的合适的基本概率分配；最后，应用经典的登普斯特组合规则将独立证据聚合成集体证据，从而对候选方案进行排序并获得最佳方案。结果：我们将我们提出的方法的结果与最近的相关工作进行比较。通过采用我们提出的方法，在示例 5 中，不确定性的置信度从 0.0751 降至 0.0051；在示例 6 中，不确定性的置信测度从 0.0547 下降到 0.0086；在示例7中，不确定性的置信测度从0.1647下降到0.0847；在应用中，不确定性的置信度从0.0069下降了0.0001。结论：提供了三个数值例子和在医学诊断中的应用，充分证明，一方面，我们提出的方法是可行和有效的；另一方面，我们提出的方法可以减少人们主观认知造成的不确定性，并以最佳性能提高选择决策水平。 (C) 2016 Elsevier B.V. 保留所有权利。
Objective: Recently, fuzzy soft sets-based decision making has attracted more and more interest. Although plenty of works have been done, they cannot provide the uncertainty or certainty of their results. To manage uncertainty is one of the most important and toughest tasks of decision making especially in medicine. In this study, we improve the performance of reducing uncertainty and raising the choice decision level in fuzzy soft set-based decision making.Methods and material: We make use of two appropriate tools (ambiguity measure and Dempster-Shafer theory of evidence) to improve fuzzy soft set-based decision making. Our proposed approach consists of three procedures: primarily, the uncertainty degree of each parameter is obtained by using ambiguity measure; next, the suitable basic probability assignment with respect to each parameter (or evidence) is constructed based on the uncertainty degree of each parameter obtained in the first step; in the end, the classical Dempster's combination rule is applied to aggregate independent evidences into the collective evidence, by which the candidate alternatives are ranked and the best alternative will be obtained.Results: We compare the results of our proposed method with the recent relative works. Through employing our presented approach, in Example 5, the belief measure of the uncertainty falls to 0.0051 from 0.0751; in Example 6, the belief measure of the uncertainty drops to 0.0086 from 0.0547; in Example 7, the belief measure of the uncertainty falls to 0.0847 from 0.1647; in application, the belief measure of the uncertainty drops 0.0001 from 0.0069.Conclusion: Three numerical examples and an application in medical diagnosis are provided to demonstrate adequately that, on the one hand, our proposed method is feasible and efficient; on the other hand, our proposed method can reduce uncertainty caused by people's subjective cognition and raise the choice decision level with the best performance. (C) 2016 Elsevier B.V. All rights reserved.