Suppliers Selection by Using a Novel Fuzzy Integral Method

[[abstract]]Supplier selection is a common problem for acquiring the necessary raw materials or services to support the business activities. However, criteria for evaluating vendors are usually complex and numerous. Furthermore, in the real life supplier selection problem, the complex criteria are usually dependent on each other, which is in contrast to the traditional assumptions on independences as well as additivity between criteria by most MCDM based provider selection papers. Few researchers tried to resolve the independence as well as additivity issue by using the Choquet integral by simply identifying a λ -fuzzy measure for the fuzzy measure identification. However, assumptions for a sole λ -fuzzy measure are usually far from realistic in the real world. Thus, this paper proposes a novel method for fuzzy measure identification based on the evaluation of pair-wise additivity degrees or interdependence coefficients between the criteria. Methods for evaluating these degrees are provided. Depending on the information being provided by decision makers on the individual density of each criterion, the fuzzy measure is constructed by solving a simple system of linear inequalities or a linear programming problem. Compared to other fuzzy measure identification approaches, the proposed method also enhances the computational efficiency, by leveraging a relatively small number of variables and constraints for linear programming. This novel method will be validated through a supplier selection problem which occurs frequently in the real world. Validation results demonstrate that the newly-proposed method can model real world MCDM problems successfully