An improved fuzzy MULTIMOORA approach for multi-criteria decision making based on objective weighting method (CCSD) and its application to technological forecasting method selection

摘要

MULTIMOORA (Multi-Objective Optimization on the basis of Ratio Analysis plus full multiplicative form) is a somewhat new multi criteria decision-making (MCDM) method which provides high efficiency and effectiveness in problem solving. To evaluate different alternatives and calculate their scores, it uses three major approaches; namely, ratio system (RS), reference point (RP), and full multiplicative form (FM). Based on the scores, alternatives are individually ranked in each approach. The obtained ranks are the basis for final ranking, which is determined under the rules of dominance theory. However, this method impose some limitations on the evaluation model such as the need for aggregation and final ranking based on each alternative's ranks already obtained by three different approaches (i.e., RS, RP, and FM) with no consideration to score differences, taking the same importance level for these approaches, the lack of attention to deviation of scores, the complexity of aggregation in the dominance theory and the need for multiple comparisons to obtain final rankings, and the probability of a circular reasoning with no distinction between alternatives included. In order to overcome shortcomings, we applied an objective weight determination method called CCSD (Correlation Coefficient and Standard Deviation) method to enhance the MULTIMOORA performance. In this regard, the scoring distance of every alternative is completely included in the aggregation. Using regression statistics, standard deviations and correlation coefficients; the dispersion in the set of scores is also computed in relation to three approaches of RS, RP, and FM. The application of unique weight for each approach will yield more realistic results. Further, it is no longer needed to use the dominance theory and the problems such as multiple comparisons and a circular reasoning are also eliminated. Finally, a real decision making problem is applied to select the proper technological forecasting method which illustrates the validity and practicality of the proposed approach.