A pattern recognition scenario, where instead of object classification into the classes by the learning set, the algorithm aims to allocate all objects to the same, the so-called "normal" class, is the research objective. Given the learning set L; the class K0 is called “normal”, and the reminder l classes K1, K2, ... , Kl from the environment K are “deviated”. The classification algorithm is for a recurrent use in a "classification, action" format. Actions Ai are defined for each “deviated” class Ki. Applied to an object x ∈ Ki, the action delivers update Ai(x) of the object. The goal is in constructing a classification algorithm A that applied repeatedly (small number of times) to the objects of L, moves the objects (correspondingly, the elements of K) to the “normal” class. In this way, the static recognition action is transferred to a dynamic domain.
This paper is continuing the discussion on the “normal” class classification problem, its theoretical postulations, possible use cases, and advantages of using logical-combinatorial approaches in solving these dynamic recognition problems. Some light relation to the topics like reinforcement learning, and recurrent neural networks are also provided.