Neural networks and neurocomputers

We consider ways to increase the intellectuality of computing systems. The theory of intellectual neural networks is presented in order to study the processes of learning, self-study and self-organization and the creation of universal artificial intelligence systems, a generalized neurocomputer model and a classification of neurocomputers by type of teaching neural networks, architecture of connections, custom crystals, reprogrammed logical integrated circuits. Particular attention is paid to the domestic neuromicoprocessors of the Neuromatrix family, the basics of programming of neurocomputers are given. The possibility of applying neuromicroprocessors as intelligent transport stream analyzers, encryption of information, image compression, radar image processing, etc