The main difficulties encountered in creating accurate fuzzy cognitive maps are presented, and a multi-stage method for adapting fuzzy relations is proposed. Preliminary results of simulation studies confirming the correctness of the proposed approach are also presented.In [9], an approach to the issue of spatial and temporal granularity of cognitive maps and certain aspects of the accuracy of such models related to it was presented. Particular attention was paid to fuzzy relational cognitive map models, described in [3-8]. A fuzzy relational cognitive map can be used to map systems with both static and dynamic structures. In the further part of the paper, the considerations will be limited to the analysis of dynamic fuzzy models.
The algorithm of operation of such models generally includes three phases:
– initial fuzzification of factor values, – information processing in the fuzzy model, – final defuzzification of factor values.
Each of the above phases can be implemented in many ways, depending on the current modeling goals [3, 7]. (…)
The article contains 16598 characters.
Source: Czasopismo Logistyka
Adaptation of relations in dynamic fuzzy relational cognitive maps – PDF file for download