DEVICE MORPHISMS IN MARKOV NETWORK. Part 2

The results of inclusion of the algorithm and Markov network with abnormal inference of a large correlator in the сNnet-scheme are presented. A new Markov network scheme is synthesized, in which the large correlator is assigned the role of a device detecting coincidences in words and compositions of interacting words – categories obtained as a result of processing input data samples from many sources of such. This new network scheme is developed for processing input data samples from organizational systems to classify the interactions of words denoting objects of these systems with some useful purpose. The main scientific goal of the study has been achieved: a method for introducing known classifiers, which are not formally considered to be constructed using the Markov algorithm, into the Markov network with abnormal inference has been substantiated and synthesized. The problem of research on the synthesis of morphisms of a large correlator in the сNnet-scheme of the Markov algorithm is solved. An additional contribution is made to solving a larger scientific problem: synthesizing a complete list of morphisms for Markov algorithms and networks. Practical examples of implementing Brouwer creativity in the control channel and by manipulating input data are considered. An additional source type associator typeн is introduced into the сNnet-scheme of the Markov algorithm, which specifies a specific method for processing input data samples for individual schemes that make up the network. In the Markov network with abnormal inference, the commands of the large correlator Dcorb for its processor Пcorb are defined. Associators of sources of a special type for controlling the large correlator are introduced into consideration and investigated. A formula of the Markov network with abnormal inference with an integrated large correlator is synthesized, it is shown that it must be developed to determine the result of executing the commands of the control channel. It was concluded that morphisms of a large correlator were synthesized and included in the сNnet-scheme of the algorithm and the Markov network of indirect propagation, which made it possible to determine the method for introducing known classifiers into it.