A Comparison of Different Normalization Method Adopted in Multi-Criteria Decision Making, Based on Some of Experimental Test Criteria’s

The main idea of this paper is to evaluate the effect of different normalization techniques in the Multi-Criteria Decision Making (MCDM) context. Hence, this paper first reviews and compares different normalization models adopted in the MCDM environment from different aspects; then, via a numerical example, the impact of different normalization methods on results is investigated. We also implicitly seek to examine the effect of normalizing the equivalent of positive and negative criteria and to analyze whether these models work the same way or not. The findings of this study show that some of the normalization techniques (i.e., the Weitendorf’s linear normalization model) are better than others. However, none of the normalization methods is perfectly complete, as trying to reach the best result in one aspect is often at the expense of achieving worse results in other aspects. In addition, the behavior of the used model for benefit-type and cost-type criteria is not the same. Finally, a new exponential version of the Weitendorf’s linear normalization model is introduced that is an improvement over the existing technique.