The paper considers issues related to the construction of a model of recognizing operators, focused on the classification of objects in conditions of high dimensionality of a feature space. A new approach to constructing a model of recognition operators based on the construction of multi-level proximity functions is proposed. In this case, the construction of the model was carried out within the framework of recognition algorithms based on the calculation of estimates. The goal of this work is develop a model is to form independent subsets of interrelated objects, highlighting a set of representative pairs of features. A distinctive feature of the proposed operator model is the definition of a suitable set of threshold functions when constructing an extremal recognition operator. The purpose of this article is to develop a model of recognizing operators based on the calculation of estimates using two-dimensional threshold rules. Scientifically, the results of this work in aggregate represent a new solution of a scientific problem related to the issues of increasing the reliability of recognizing operators based on the evaluation of estimates. The practical significance of the results lies in the fact that the developed operators and programs can be applied in medical and technical diagnostics, geological forecasting, biometric identification and other areas where it is possible to solve the problem of classifying objects defined in a space of large dimensionality.

The problems of determining the representative features when constructing an extremal pattern recognition algorithm defined in a high-dimensional feature space are considered. As an initial model, a model of recognition algorithms based on radial functions was considered. The procedures for extracting subsets of interrelated features and selecting a set of representative features when constructing an extremal algorithm within the framework of the recognition model under consideration are presented. The main idea of the first procedure is to form a set of features that are unrelated (or are connected rather weakly) among themselves. The essence of it is as follows. The studied subsets of features are combined into one subset if they are close enough to each other in a sense. Otherwise, they belong to different subsets. The main idea of the second procedure is to define a representative element in each subset of tightly coupled features. The choice of an element from the considered subset of tightly coupled features is made on the basis of an assessment of the proximity of features. It is required that among the selected set of features there should not be tightly coupled features. The main advantage of the proposed procedures is to improve the accuracy of the results of allocating subsets of tightly coupled features when building a recognition algorithm under conditions of large dimensionality of the feature space and determining the quantitative evaluation of these subsets. In order to assess the performance of the developed procedures, experimental studies were carried out. The application of the developed procedures allows to more accurately determine the unknown parameters of recognition operators in the space of features of large dimension.

One of the urgent problems in the field of computer vision is the problem of developing algorithms for recognizing objects represented as images. To solve the problem under consideration, recognition algorithms based on radial functions are proposed. The main idea of the proposed algorithms is to form a set of preferred features that simplify the process of developing an extreme algorithm when recognizing images represented as images. To test the efficiency of the proposed algorithms, the problem of diagnostics of diseases of agricultural crops, in particular cotton, according to the images of their leaves, is solved. The results of the conducted experiments show that the application of the proposed algorithms allows increasing the accuracy in solving the problem of cotton diseases diagnosing on the images of its leaves.

The problem of constructing a model of recognition algorithms in the conditions of high dimensionality of feature space is considered. To solve this problem, a new approach suggests that it takes into account the interconnectedness of the given features. The main concern of this approach is to isolate representative traits and form preferred combinations of traits. A distinctive feature of the proposed model of algorithms is to determine a suitable set of elementary threshold rules within the framework of the preferred combinations of representative features and build a proximity function based on this set of feature combinations. Scientifically, the results of this work in the aggregate represent a new solution to a scientific problem related to improving the reliability of recognition algorithms based on radial functions. The practical significance of the results lies in the fact that the developed algorithms can be applied in medical and technical diagnostics, geological forecasting, biometric identification. To test the performance of the proposed model, experimental studies were performed.