一种基于邻域粗糙集和TOPSIS法的组合决策方法及应用

A combination decision method based on neighborhood rough set and TOPSIS method and its application

传统的粗糙集-TOPSIS组合决策方法有效解决了TOPSIS法在构造评价矩阵时存在主观性的问题,提高了决策精度。但该方法只适应于处理离散型数据的决策表信息系统,对于连续数据的决策表信息系统,需要对数据进行离散化处理,而离散化处理会改变数据原始属性,产生数据失真,从而导致决策精度显著下降。针对此类问题,结合邻域粗糙集理论,提出了一种邻域分辨矩阵改进的邻域粗糙集-TOPSIS综合评价方法。该方法可以直接应用于连续数据的决策表信息系统,不仅能计算核心评价指标权重的重要度,而且能计算非核心评价指标权重的重要度;在辐射源威胁排序案例中的应用表明,其可显著提高连续数据决策表信息系统的决策精度。

The traditional rough set TOPSIS combination decision-making method effectively solved the subjectivity problem in constructing evaluation matrices with TOPSIS method,and improved decision-making accuracy.However,this method is only applicable to decision table information systems that handle discrete data.For decision table information systems with continuous data,discretization of the data is required,which can change the original attributes of the data and cause data distortion,resulting in a significant decrease in decision accuracy.In response to such problems,a neighborhood resolution matrix improved neighborhood rough set TOPSIS comprehensive evaluation method is proposed by combining neighborhood rough set theory.This method can directly process decision table information systems with continuous data,not only calculating the importance of core evaluation index weights,but also calculating the importance of non core evaluation index weights.The application in radiation source threat ranking cases shows that it can significantly improve the decision-making accuracy of continuous data decision table information systems.