Using Set Covering Approach for Decision-Making Criteria Classification while Correlation between Criteria Exist

In choosing an alternative among many, decision makers usually take into account multiple criteria’s, which are sometimes in line with each other and sometimes conflicting. When there is dependencies between criteria, MCDM techniques such as ANP, AHP and DEMATEL could be applied. Another approach is to classify criteria that are correlated with each other. In this study a new approach of set covering problem to this end has been used. Our objective is setting these criteria into clusters so that their correlation in each cluster to be maximized, and also their correlation with others be minimized. In order to validate our proposed model, factor analysis method used. Data gathered from 9 industries related to 26 foreign countries. Finally, results of both models were almost similar. The results show that efficiency of the proposed method (set covering) for grouping variables is acceptable and this method in future research can be used to classify variables (when there is a correlation between variables).