一种基于偏差熵的组合评价方法及其应用

传统组合评价方法的一般思路是使组合评价结果与基于各单一评价方法所得结果之间的偏差尽可能小,但该思路对相容方法集本身的质量却考虑较少。针对上述组合评价思路的不足,提出了偏差熵的概念,并在此基础上提出了基于偏差熵的组合评价方法。该方法对传统组合评价方法所做的改进在于,在考虑减小偏差的同时亦考虑了相容方法集本身的相容质量问题。最后,通过算例验证了用该方法得出的评价结果更加全面、合理,且该方法易于实现。

Combination Evaluation Method Based on Deviation Entropy and Its Application

The general research idea of traditional combined evaluation method is to make the deviation between combined evaluation result and results obtained by single evaluation method as small as possible,but this idea considers little the quality of compatible method set.To solve this problem,this paper puts forward the concept of deviation entropy,and proposes a combined evaluation method based on this concept,which improves the traditional combined evaluation method through considering the compatibility quality of compatible method set and decreasing deviation.Finally,it demonstrates the effectiveness of this proposed method through a numerical example.