In many areas, the criteria for the adoption of certain decisions are not quantitative, but qualitative assessments of some crucial criteria. This applies, in particular, to the integral assessment of the health status of patients, or the assessment of the ecological state of the environment, etc. Therefore, an integral qualitative assessment of an object, based on the known numerical states of its individual elements, is an urgent non-trivial task. Currently, there are methods for calculating the quantitative integral assessment, which is a kind of “code” for the qualitative assessment. A direct verbal evaluation is obtained, as a rule, by ranking the possible numerical values of the resulting integral code, and assigning to each given interval a certain qualitative definition. However, well-known methods for constructing analytical computational integrated assessment are not effective. As an alternative, the authors of the article propose an approach consisting in a comprehensive assessment of the state of objects based on the allocation of “similar” groups, and analysis of the basic general properties of objects in the group. Such problems can be solved by cluster analysis methods. Cluster analysis allows you to group (decompose) data that has the property of “similarity” in a given sense and to separate them from “dissimilar” data. By analyzing data in groups, experts can give a high-quality interpretation of homogeneous data in a cluster. However, when analyzing the obtained groupings by domain experts, a situation often arises when the data in some groups have a sufficient degree of homogeneity in order to give them a qualitative assessment, and in others the mathematical method of clustering does not allow to separate heterogeneous data from each other, and experts cannot give rating. So the idea to apply step-by-step clusterization when in each step it is decided whether every cluster has a sufficient degree of homogeneity or not. If not, all non-homogeneous clusters should be further decomposed into two or more clusters. It remained to decide which clustering method to choose for the data decomposition. To address this problem, reconnaissance experiments were conducted. As a result, the Kohonen Self-Organizing Map (SOM-cards) method has been proved to be the best. The proposed algorithm for multilevel clustering was called the Cascade Neural Network Filtering Data. Its effectiveness was confirmed by numerical experiments.

The article describes decision-making methods based on intelligent learning algorithms, for the construction of which verbal elements are used. Such algorithms and methods operate in calculations with strictly quantitative data, however, taking into account the human way of perceiving information in verbal form. A person does not directly participate in the process of constructing a model, that is, its structure does not depend on expert or other human opinions, however, qualitative verbal information (for example, elements of regulatory acts, documents, orders, etc.) is embedded in the algorithm in encoded form. Unlike the classical methods of fuzzy logic and fuzzy inference, the proposed models consist of several interconnected classical and author’s models, using author’s numerical methods. This increases their accuracy and adequacy. The great advantage is the automatic way to build these models, only on the basis of sets of initial quantitative and qualitative data, which will allow you to quickly and efficiently create recommender systems for solving a wide range of decision-making problems. The system is built in three stages: First, the fuzzy logical inference Takagi – Sugeno – Kang system (TSK) is automatically built on the sounding of the available quantitative and qualitative data. The advantage of such a system is that it can be retrained (adjust its parameters). Then, after the system is built according to the available data, at the second stage its accuracy is increased using a fuzzy neural network. The drawback of the constructed system is that it does not give out a verbal answer. At the third stage, based on the adjusted parameters of the TSK system, a Mamdani type system is constructed using a special algorithm. As a result, it can produce a verbal answer from a recommender system, and at the same time gets all the advantages of a neural network refinement. Computational experiments are presented.

The article describes a method for the phased finding of the multidimensional object class in the case when the set of classes is not known in advance. The developed method first solves the problem of selecting classes from an untyped heterogeneous set of objects, and then classifies an arbitrary new object into the determined classes. Classes are found on the basis of the author's algorithm of cascade neural filtering, and the objects classification is performed using the author's model based on a finite automaton.

No abstract