考虑专家判断信息的灰色关联极大熵权重模型

本文提出了一种从关联角度出发将主观先验信息与客观信息纳入约束条件从而求解综合权重的方法。在利用灰关联深度系数对实际决策问题进行客观权重判断研究的基础之上,构建了源于专家判断信息的权重势比的主观约束条件,将主客观因素同时反映在优化模型的约束条件中,并将权重的极大熵作为目标函数,保证权重判断的可信度,从而构建了确定评价指标综合权重的极大熵优化模型。该方法克服了将主客观条件直接通过线性组合作为目标函数时,主客观参数选取导致的权重大小的不确定性,并同其他赋权方法进行案例结果比较,表明了该方法的有效性。

The Maximum Entropy Empowerment Model for Evaluating Index Considering the Expert Evaluation Information

For comprehensive determining indexes’ weight of the multi-attribute decision-making,the article proposes a method considering prior subjective information and objective information into constraints from the angle of the association to estimate the indexes’ synthesis weight.Based on solving the objective weight using the grey correlation deep coefficient on the actual decision problem,this paper focuses on building the subjective constraint condition constructed with the potential ratio of index weight confirmed by specialist which confirm that the subjective and objective factors can be reflected simultaneously in the constraint conditions of the optimization model.The maximum entropy optimization model is built to ensure the credibility of the weight judgment,thus establishing the maximum entropy optimization model for synthesis weights.This method overcomes the uncertainty of weight because of the subjective and objective parameter selection when the subjective and objective conditions are directly grouped through the linear objective function.Finally,comparison with another methods based on a numerical example shows a better effectiveness and practicability of this method.