当前位置: X-MOL 学术Int. J. Comput. Sci. Eng. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Dealing with value constraints in decision making using MCDM methods
Journal of Computational Science ( IF 3.3 ) Pub Date : 2020-06-13 , DOI: 10.1016/j.jocs.2020.101154
Abdelkrim Abdelli , Lynda Mokdad , Youcef Hammal

In our daily life, we are always facing decision making problems with different complexities and requirements. Therefore, the need to design new theories, methods and tools to solve these kinds of problems, as efficiently as possible, becomes a real challenge.

In the current research, we deal with decision problems wherein value constraints are expressed on the performance ratings of the alternatives. We focus on MCDM (Multi-Criteria Decision Making) methods to solve such problems. When value constraints are specified, traditional MCDM methods proceed upstream by removing from the research space all the non-satisfactory alternatives, while more recent approaches provide a subjective ranking of these alternatives. To overcome some of these issues, we have introduced, recently, a new MCDM method called, ISOCOV (Ideal Solution with Constraint on Values). ISOCOV aims at providing to the decision maker a more accurate solution while dealing with the given value constraints.

We propose in this paper to introduce more flexibility in this method and study its accuracy by comparing its rendering with other methods through its application on a real dataset. The adaptation introduced in our study makes it possible to specify the problem by assuming the value constraints either as mandatory (hard constraints), or as non-compulsory (soft constraints). In the first case, the alternatives that are meeting all the constraints are ranked on the top, whereas the rest of the field is ranked below. However, if the constraints are soft, ISOCOV ranks the alternatives according to the combination of their performance ratings and their closeness to meet all the constraints.



中文翻译:

使用MCDM方法处理决策中的价值约束

在我们的日常生活中,我们始终面临着具有不同复杂性和要求的决策问题。因此,需要设计新的理论,方法和工具以尽可能有效地解决这类问题,这是一个真正的挑战。

在当前的研究中,我们处理决策问题,其中价值约束表示在替代方案的性能等级上。我们专注于解决此类问题的MCDM(多标准决策)方法。当指定了价值约束时,传统的MCDM方法通过从研究领域中删除所有不令人满意的替代方案而向上游进行,而最新的方法则提供了这些替代方案的主观排名。为了克服其中的一些问题,我们最近引入了一种新的MCDM方法,称为ISOCOV(具有值约束的理想解决方案)。ISOCOV旨在为决策者提供更准确的解决方案,同时处理给定的价值约束。

我们建议在本文中介绍此方法的更多灵活性,并通过将其渲染与其他方法进行比较(通过将其应用于实际数据集)来研究其准确性。在我们的研究中引入的适应性可以通过将值约束假定为强制性(硬约束)或非强制性(软约束)来指定问题。在第一种情况下,满足所有约束的替代方案排在顶部,而其余字段排在下方。但是,如果约束条件较弱,则ISOCOV会根据其性能等级和满足所有约束条件的接近程度的组合来对替代方案进行排名。

更新日期:2020-06-13
down
wechat
bug