Fuzzy Graphs Clustering with Quality Relation Functionals in Cognitive Models

In this study we present a new approach for developing input-output data set (antecedents, consequents) for fuzzy rules of expert system production based on the mechanism of fuzzy logic inference. Integration of the methods for cognitive modeling and of analysis with the expert system has been proposed. To generate the logical conclusion attributes on the basis of fuzzy graph models, the clustering procedure and the detection of system response are used. The approach, called Self-Constructing Attribute Generator, SCAG, consists in consecutive transformation of the initial matrix of the fuzzy graph model using two types of quality functionals. The first step is the initialization of the primary transformation matrix to upper-triangular sight using the square barrier penalty functions and “inverse” functions. At the second stage, the feedback on disturbances is generated in the form of a vector set. Further graph clustering is directly made based on the minimization of the potential energy functional.