Симуляторы нейронных сетей

This paper presents the functional model of signal processing in networks of similar interconnected active elements. On the one hand, it is a macroscopic model of the dynamic interactions between neural networks and its states can be considered as normal and abnormal modes of signal processing (for example, images). On the other hand, the model and the calculation results can be considered as a variant of modeling epilepsy. In this paper we develop an information macroscopic approach to the modeling of epilepsy, which in this sense is unique.

Now models (simulators) of neural networks are actively developed. Their architecture and design are based on features of structure and principles of work of real neurons and neurobiological systems. Working out neurolike models based on the data about architecture of connections in a brain, it is aimed at finding­out of principles of work of its neural structures. In experimental researches it is revealed that interconnected neuronal modules such as cortex, reticular modules of thalamus, specific thalamus play the important role in processes of information processing.