信息熵和判断矩阵的专家聚类赋权法

Expert Cluster Weighting Method Based on the Information Entropy and the Judge Matrix

为了降低专家聚类赋权过程中由排序向量引起的不确定性和判断矩阵引起的矛盾性,提高权值分配的精确性,提出了基于信息熵和判断矩阵相结合的专家聚类赋权法。该方法采用聚类分析原理,对排序向量进行分类,根据分类结果、信息熵值和一致性比率确定专家权重系数。实例分析表明:基于信息熵和判断矩阵相结合的专家聚类赋权法在具体应用中得到的结果离期望值更近,说明该算法有效可行。

In order to reduce the uncertainty caused by sorting vector and the contradiction caused by judgment matrix in the process of expert cluster weighting method,improve the accuracy of weights allocation,an expert cluster weighting method is proposed based on the combination of information entropy and judgment matrix.This method uses the clustering analysis principle to classify scheduling vector,then the weight coefficient is determined according to the classified results and information entropy and consistency ratio.Example shows that： experts clustering weighting method based on the information entropy and the judgment matrix is closer to the expectations in specific application,so this algorithm is effective and feasible.