Method for Classifying Economic Agents Based on Neural Networks and Fuzzy Logic

The paper is devoted to the creating problem a method for classifying economic agents considering fuzzy information. As possible economic agents that are subject to classification, there may be borrowers, audit objects, stock buffers, and others. The novelty of the study lies in the fact that the proposed method of neuro-fuzzy classification provides the representation of knowledge about the classification of objects in the form of fuzzy rules that are convenient for perception; does not require the involvement of experts whose knowledge is difficult to extract and which may be incomplete; can be used in situations where the number of fuzzy values of a linguistic variable is large and their description by linguistic constructions is difficult; reduces computational complexity, standard error and classification probability by automatically selecting the model structure and parameters. As a result of a numerical study, it was found that the proposed classification method based on neural networks and fuzzy logic provides a classification probability of 0.98, and an MSЕ of 0.05. Further research prospects are the use of the proposed classification method based on neural networks and fuzzy logic for various intelligent object classification systems.