Regularized Identification of Impulse Processes in Cognitive Maps with Structural Uncertainty

Dynamics of complex systems of different nature is described by difference mathematical models of impulse processes in cognitive maps (CM). The nature and sense of some CM nodes can be established based on system analysis, but other nodes coordinates influencing the system remain apriori uncertain due to different reasons. Hidden internal relations in the complex system results in uncertainty of its representation as a directed graph (CM), i.e., in unknown CM dimension. Moreover, if supposed CM dimension appears to be high then errors in measurements nodes and uncontrolled disturbances usually cause identification problem to be ill-posed. The present report gives regularized solution of the identification problem inCM impulse processes where the CM dimension stands for regularization parameter.