RETRACTED CHAPTER: Towards the Convergence in Fuzzy Cognitive Maps Based Decision-Making Models
Roughly speaking, decision-making can be defined as the process to select a decision (or group of decisions) among a set of possible alternatives in a given decision activity. Most real-life problems are unstructured in nature, often involving vagueness and uncertainty. This makes difficult to apply exact models, being necessary to use approximate methods based on Soft Computing techniques. In recent years, Fuzzy Cognitive Maps have been used in designing Decision Support Systems due to their capability for explaining the underlying reasoning process. This includes the development of learning methodologies for adjusting the inherent parametric requirements. Less attention has been given to the map convergence and its implications in the decision process. In this paper, we study the convergence issues of Fuzzy Cognitive Map based models used in decision-making. More explicitly, we present a learning procedure that allows improving the network convergence by preserving the ordinal relation between the alternatives. In this learning algorithm, the direction and intensity of causal relations cannot be altered since they comprise the system semantic. Numerical simulations show the practical usability of theoretical contributions proposed in this paper, when solving decision-making problems.