众里取大规则下由频率确定属性权重的方法

发布时间：2018-04-12 16:40

本文选题：多属性决策 + 权重　；参考：《统计与决策》2017年08期

【摘要】：目前获取属性权重方法,大多数需要决策者对属性直接打分或两两比较或优劣排序,这要求属性数量不宜过多、决策者应对各个属性有充分了解。众里取大规则下由频率确定属性权重的方法要求决策者挑选一个最重要属性,进而通过频率确定属性权重。文章首先视属性赋值为随机变量,依概率期望定义权重公式;然后根据离散选择模型相关结论,推导出属性频率与权重的转换表达式,将其代入权重公式得到属性权重;最后构造卡方统计量,判断权重之间是否具有显著差异。该方法适用于有大量决策者和众多属性的群决策问题。
[Abstract]:At present, most of the methods of obtaining attribute weight require the decision maker to rate the attributes directly or compare them or rank them, which requires that the number of attributes should not be excessive, and the decision makers should have a good understanding of each attribute.The method of determining attribute weight by frequency under the rule of mass selection requires the decision maker to select one of the most important attributes and then determine the attribute weight by frequency.In this paper, the attribute assignment is regarded as a random variable, and the weight formula is defined according to the probability expectation. Then, according to the relevant conclusions of the discrete selection model, the conversion expression of attribute frequency and weight is deduced, and the attribute weight is obtained by replacing it with the weight formula.Finally, chi-square statistics are constructed to determine whether there are significant differences between weights.This method is suitable for group decision making problems with a large number of decision makers and many attributes.
【作者单位】：辽宁工程技术大学公共管理与法学院;黑河学院经济管理学院;
【基金】：辽宁省社会科学规划基金合作项目(L15EJY001) 辽宁省教育厅项目(LJCR007;LGCR010)
【分类号】：C934

【相似文献】