Fuzzy TODIM method based on alpha-level sets

Abstract Fuzzy TODIM method has been widely and successfully used to solve different decision making problems and consider decision maker’s (DM’s) psychological behavior during the decision process under uncertain environment. In the existing fuzzy TODIM methods, the fuzzy preferences are defuzzified into crisp values by using a distance measure between fuzzy numbers or variables. However, such an operation may imply a significant loss of information. It is argued that if fuzzy information is defuzzified into crisp values at the very beginning, then the superiority of considering fuzzy information is discounted. Therefore, it seems better to keep as much information as possible during the decision process rather than oversimplifying the fuzzy information by crisp values. An effective and proper way for keeping as much information as possible dealing with fuzzy preferences during the decision process is the use of alpha level sets. Several fuzzy multi-criteria decision making (MCDM) methods based on alpha level sets have been proposed and used to handle fuzzy information successfully, however, they neglected DM’s psychological behavior that plays a critical role in the real world decision processes. Up to now, there is not a fuzzy MCDM method employing alpha level sets to deal with fuzzy information together with considering DM’s psychological behavior. Motivated by previous limitations, this study proposes a novel fuzzy TODIM method based on alpha level sets, which keeps the fuzzy information longer and considers DM’s psychological behavior in the decision process. In addition, different ways to select the best alternative are provided in the proposed method. Comparisons with several MCDM methods are presented to show the improvements both of dealing with alpha level sets when psychological behavior is considered and the sensitivity of using such behavior regarding those MCDM methods that do not consider it. Through the comparison analysis, the proposed method is significant superiority to the existing approaches, which not only improves the current studies, but also enriches the way of coping with fuzzy information in the extant fuzzy MCDM methods.