论文部分内容阅读
提出了一种基于粗糙集分类来判断群决策专家个体可信度权重的新方法。先根据专家给出的判断矩阵分别按照条件属性值和决策属性值将方案进行分类,再用概率分布函数来考察每个条件属性类在决策属性类中的分布情况,算出与均匀分布时的概率分布函数的距离和,将各个专家判断矩阵的距离和进行归一化处理,以此作为群决策专家个体可信度的权值;最后用算例给出了该方法的可行性和有效性。
A new method for judging the weight of individuals in group decision-making experts based on rough set classification is proposed. According to the judgment matrix given by experts, the schemes are classified respectively according to the condition attribute values and the decision attribute values, and then the probability distribution function is used to examine the distribution of each condition attribute class in the decision attribute class to calculate the probability of being uniformly distributed The distance of distribution function and the distance of each expert judgment matrix are normalized, which is used as the weight of individual credibility of group decision-making experts. Finally, an example is given to show the feasibility and effectiveness of the method.