A generalized TOPSIS method for group decision making with heterogeneous information in a dynamic environment
Rodolfo Lourenzutti,Renato A. Krohling
DOI: https://doi.org/10.1016/j.ins.2015.10.005
IF: 8.1
2016-02-01
Information Sciences
Abstract:In the real world the decisions are frequently made by a group of decision makers. Methods to support group multicriteria decision making (MCDM) in dynamic environments is a challenging research topic under investigation. However, in most of those methods, it is necessary that the decision makers reach an agreement in the setup of the problem. For example, it is common that a group MCDM method requires the decision makers to define jointly a set of criteria. This may not be easy or, even, achievable. Also, the MCDM methods have been extensively generalized to process many different types of information, e.g., crisp, interval, fuzzy, intuitionistic fuzzy, hesitant fuzzy. Nevertheless, many group MCDM methods strongly restrict the freedom of the decision makers to use the type of information they see fit by forcing them to prior define the type of information that must be used. These restrictions considerably reduce the individual opinions of the decision makers involved. In this work, we introduce a generalization of the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method, called GMo-RTOPSIS (Group Modular Random TOPSIS), which provides freedom for the decision makers express his/her individuality and opinions. The method is capable of dealing with an imperfect setting where each decision maker can define independently the criteria set, the weight vector, the underlying factors that may affect the alternatives’ ratings and the type of information they want to use in each criterion. We then show the feasibility of the method by discussing three case studies.
computer science, information systems