Construction of a Logical-Algebraic Corrector for Increasing Adaptive Properties of the $\Sigma\Pi$-Neuron

In this paper, we consider the problem of constructing a correction algorithm for increasing adaptive properties of the $\Sigma\Pi$-neuron, based solely on the structure of the $\Sigma\Pi$-neuron itself. The logical-algebraic method of data analysis is used for the construction of the corrector. Comparison of advantages of the neural-network approach and the logical-algebraic method leads to the conclusion that the combined approach to the organization of neural networks improves their efficiency and allows one to state rules that reveal hidden patterns in a given subject area and thus to improve the quality of the recognition system.