A Structural Expansion Method of Fuzzy Cognitive Map Under Cognitive Limitations

With the diversification of social development, the difficulty and complexity of prediction problems have increased significantly. However, most prediction models are not interpretative, can only deal with limited and relatively simple problems, and lack solutions to complex problems. As a data-driven discrete nonlinear system, Fuzzy Cognitive Map combines the advantages of neural network and fuzzy logic, has strong knowledge representation and logical reasoning ability. However, the nodes selected in complex systems cannot be comprehensive and practical due to the limitations of human cognition, and the lack of essential nodes will affect the accuracy of the prediction model. This paper will focus on the structure of FCM and analyze the FCM prediction model under cognitive limitations: The results show that cognitive limitations reduce the prediction accuracy; We propose a complete cognitive structure fuzzy cognitive map method (CC-FCM) to improve the prediction accuracy and solve the modeling and prediction problems of complex systems.