Identification of telltale signature by using method of adaptive neuro-fuzzy systems

Abstract

The article is devoted to the problem of using the method of adaptive neuro-fuzzy networks to assess the informativeness of the telltale signature of the object of protection. In ANFIS, a fuzzy system is used to represent knowledge in an interpreted form, as well as the ability to train a neural network to optimize the parameters of a fuzzy system. The training data base was based on the results of the previous study, which consisted in building a decision support subsystem for identifying the set of telltale signature of a protected object based on expert judgment. It contains information about the informativeness of each group of telltale signature and is used in the process of learning a neural network. Constantly changing the parameters of the membership functions when learning a neuro-fuzzy system allows you to more accurately configure the system to solve the task. The study showed the possibility of using the method of adaptive neuro-fuzzy networks in the process of evaluating the informativity of the telltale signature.