基于相对熵的多属性决策排序方法

针对逼近理想解排序法(TOPSIS)的不足,从相对熵的概念出发,提出了求解多属性决策问题的新思路.利用被评价方案与理想方案和负理想方案的相对熵,定义了一种新的与理想方案的贴近度.据此给出一种新的排序方法--相对熵排序法,并将此法与TOPSIS方法、夹角度量法和投影法进行对比.结果表明,在传统的TOPSIS方法、夹角度量法和投影算法排序都失效的情况下,相对熵排序法能够准确地给出方案的排序.

Relative entropy evaluation method for multiple attribute decision making

To overcome the shortcomings of the technique for order preferenceby similarity to ideal(TOPSIS), this paper
presents a new method to solve the problem in the multiple attribute decision making based on relative entropy. By using
the relative entropy between an alternative and the ideal solution, and between the alternative and the negative-ideal solution,
a new relative closeness to the ideal solutions is defined. Therefore, a new synthetically evaluation method-relative entropy
evaluation method is developed. This method is compared with TOPSIS method, angle measure evaluation method and
project method, which indicates that the relative entropy evaluation method can get good results when other methods all fail.