PROTOMORPHISMS IN THE THEORY OF MARKOV ALGORITHMS AND BIG DATA OF SOCIO-ECONOMIC SYSTEMS

Abstract

Protomorphisms known in category theory are investigated together with the theory of Markov algorithms. Protomorphisms are considered as an element extending the theoretical provisions of the theory of Markov algorithms. Protomorphisms are investigated in the big data of socio-economic systems in the tasks of countering money laundering and terrorist financing. Methods of assignment for protomorphisms of the absence of a preorder are considered. A synthesized method for recording protomorphism in a data cycle when entering input sample sets into a big data classification algorithm is considered. A method of writing a protomorphism using the alphabet M = λιτφп/αβδγχκμ for the N-scheme of the Markov algorithm is presented, by entering a new nonterminal symbol into it. Examples of protomorphisms in big data are considered. Protomorphisms are divided into those about the truth, which can be asked, and those for which it is impractical to make a judgment about truth or falsity. The Hempel paradox with protomorphisms is considered. For Hempel’s paradox with “classes”, the method of bringing the number of inductive observations to the total number of observations is specified. An example of filling an artificial intelligence database with protomorphism is considered. A practical example of the use of protomorphism in the task of countering money laundering and terrorist financing of the socio-economic system is given. It is shown that it is permissible to expand the alphabet M with new letters for some morphisms, separating them. It is concluded that in the task of countering money laundering and terrorist financing, it is advisable to use protomorphisms, since they can increase the level of completeness, reliability, and efficiency of the subsystem for countering money laundering and terrorist financing when making decisions.