| 摘要: |
| 传统TOPSIS评价法构造的正、负理想解容易产生逆序现象，而且使用的欧氏距离无法考虑到决策指标之间的相关性；针对此类问题，提出了一种基于加权马氏距离的TOPSIS改进法：一是用极差变换法对评价指标进行无纲化处理，消除逆序现象；二是对评价方案做正交变换，利用变换后的评价方案到正、负理想解的马氏距离代替传统欧式距离，经案例证明，改进后的TOPSIS优于传统的方法。 |
| 关键词: 逆序 加权马氏距离 协方差阵 TOPSIS |
| DOI： |
| 分类号: |
| 基金项目: |
|
| An Improved TOPSIS Evaluation Method Based on Mahalanohis Distance and Its Application |
|
WANG Wei xia
|
| Abstract: |
| Positive and negative ideal solution created by traditional TOPSIS evaluation method is easy to cause reversal problem and the used Euclidean distance can not consider the correlation between decision making indicators.For such problems，this paper proposes an improved TOPSIS method based on weighted Mahalanohis distance.The first is using the maximum difference dormalization method to eliminate the reversal problem in the traditional TOPSIS method，and the second is using the orthogonal transformation method by using the evaluation scheme after the transformation to calculate Mahalanohis distance to positive and negative ideal solution to replace the traditional Euclidean distance.The case analysis proves that the improved TOPSIS method is superior to the traditional method. |
| Key words: reversal problem weighted Mahalanohis distance covariance matrix TOPSIS |
