Different dynamic causal relationship approaches for cognitive maps

We study in this work the problem of adaptation on cognitive maps (CMs). We review different approaches of adaptation for CM, based on the idea that the causal relationships of the CM change during their phase of execution (runtime). Particularly, we study three dynamic causal relationships: the first one where the relationships between the concepts are defined as fuzzy rules, and the concepts and the relationship are fuzzy variables; the second one where mathematical models that describe the real system are used to define the causal relationships; and finally, in the last one the causal relationships are defined by generic logic rules based on the state of the concepts of the map. Each one can be used to model different types of systems, because each one exploits specific characteristics of the modeled system. These approaches are tested in different problems, giving very good results, and demonstrating that the utilization of CM as dynamic models is reliable and good.