基于Spearman秩相关系数的正态模糊数决策方法

The Decision-making Method of Normal Fuzzy Numbers Based on Spearman Rank Correlation Coefficient

针对属性值为正态模糊数,属性权重未知但是权重秩次已知的多属性决策问题,考虑到属性间的重要性差异,提出一种以Spearman秩相关系数为测度指标的决策方法。首先利用正态模糊数相似度得到衡量方案和理想方案的相似度向量。其次利用Spearman秩相关系数度量方案的相似度向量和属性权重向量间的秩次相关性,以此作为判断方案优劣的依据。最后通过实例说明该方法的合理性和有效性。

For the multi-attribute decision problem which the attribute value is given in the form of normal fuzzy numbers and the attribute weight is unknown but only the weight rank order information is known,and considering the differences in the importance of each attribute indicator,the decision method is proposed based on Spearman rank correlation coefficient as measurement index.Firstly,the similarity of normal fuzzy numbers is used to calculate the similarity vector of the scheme and the ideal scheme.Secondly,Spearman rank correlation coefficient is used to measure the rank correlation between the similarity vector and the attribute weight vector,which is used as a criterion to determine the merits of the scheme.Finally,an example is given to show the rationality and effectiveness of the method.