Introduction

The concept of stratification (CST) is an innovative approach to problem solving which has recently been proposed by Zadeh (2016). CST is a system that receives inputs which are the basis for transitioning through different states (Asadabadi, Saberi, & Chang, 2017; Asadabadi, Saberi, & Chang, 2018a, 2018b). In each state the inputs are coupled with outputs and this enables the structuring of dynamic situations such as environments in which multiple criteria decisions are made. Various Multiple Criteria Decision Making (MCDM) methods have previously been developed (Govindan, Rajendran, Sarkis, & Murugesan, 2015) and used where a number of alternatives need to be ranked based on selection criteria. However, it often occurs that the decision maker is doubtful about the final decision when an MCDM method is employed (Diaz-Balteiro, GonzálezPachón, & Romero, 2017). Such doubt is due to the fact that the future is always accompanied by uncertainty and the uncertainty makes the decision maker doubtful about the weights assigned to the criteria (Asadabadi, 2017). There are many cases, such as the case discussed in Section 4 of this paper, in which the decision maker is not quite certain when static weightings are provided for the criteria to be used in MCDM methods. So far, MCDM methods have not been empowered to consider fluctuations, in the weightings of decision criteria, in the way changes occur in the human brain. Reviewing the literature (in Section 2) we notice that there is a research gap and this problem has not been sufficiently investigated. Developing a method to resolve doubt in the decision-making process has motivated this study. Such a method can be developed by utilising supportive concepts such as CST for considering changes that are likely to happen in the decision environment. While making a decision, observing the decision environment and anticipating and considering possible situations can increase the robustness of the final decision. Such an approach to decision-making mimics the decision processes of the human brain while eliminating the associated confusion. When the human brain takes multiple criteria into account in order to make the best decision, the brain considers many positive and negative situations that might happen (Steyvers, Lee, & Wagenmakers, 2009; Weng, Huang, & Li, 2010). This means frequent changes in the weightings of the criteria. For example, when deciding to rent a unit several ‘what ifs’ come to mind: the possibility of having guests, having kids, buying a car or getting a new job. Such uncertainties can change the importance weightings of the criteria such as price, distance and size of the unit. As the importance weightings start changing, the relative value of different units/alternatives may change, and this may impact the decision. Since the human brain is not capable of considering all of the relevant situations simultaneously (Tzeng & Huang, 2011), the decision maker might be confused about whether the right decision is being made. The contribution of this paper is in providing an application of a recent concept, namely CST, and in showing how this concept can be utilised in combination with an MCDM method to structure the decision-making process in a way that is similar to what takes place in the human brain. In doing so, different eventualities are taken into account while making a multi-criteria decision. In the proposed method, the current state of a decision is identified and potential states that may occur, and are adjacent to the current state, are also engaged in making the decision. This consideration strengthens and empowers MCDM methods by enabling them to handle the dynamicity of the decision environment. A process that stratifies the environment and the associated precomputations benefits the decision maker by ensuring them that their concerns are taken into account in the decision process. As a result, there is less likelihood of regret in the future. The proposed integration should stimulate work on the future application of CST in conjunction with various MCDM methods, in particular, in artificial intelligence. The remainder of this paper is structured as follows. The next section features a brief review of the current literature on MCDM methods and CST.