Vol 46, No 6 (2019)

This article deals with the problem of self-consciousness modeling for intelligent agents with a sign world view. A conceptual model of self-consciousness is proposed, which includes a description of the “I” sign components and the typology of its relations with other signs of the world view. The empirical research of the “I” sign semantic component was conducted using the “RSA machine,” which enables a deep linguistic analysis in the interests of socio-humanitarian sciences, and the results of the research are presented. It is concluded that a certain role behavior (e.g., a victim or a successful self-manager) affects a person’s unconscious use of predicates with some semantics.

This paper considers a new approach for designing fuzzy rule-based classifiers. To optimize the parameters of classifiers, a continuous shuffled frog-leaping algorithm is applied. On a set of constructed classifiers, the optimal classifier is selected in terms of the accuracy and the number of features used, using the statistical Akaike informational criterion. The efficiency of the proposed approach is tested on 15 KEEL data sets. The results are statistically compared with the results of similar algorithms. The new approach to designing fuzzy classifiers proposed in this article makes it possible to reduce the number of rules and attributes, thereby increasing the interpretability of classification results.

A problem of multicriteria optimization on a fuzzy set specified by its membership function is considered. A two-step approach is proposed to solve this problem. It involves searching for a compromise between the available criteria and the membership function. At the first stage, a new vector criterion is formed by adding the membership function to the set of initial criteria and information about the importance of criteria in the form of quanta of information is used to reduce the Pareto set. If the reduced Pareto set is not chosen as the final solution to the multicriteria optimization problem, scalarization that uses the concept of ​​goal programming is proposed at the second stage.

An algorithm for solving the problem of analysis of connectivity of a map compiled by the distributed information robot system is described. This task can be implemented by a flock (or a swarm, which is a flock without a leader) of flying reconnaissance robots, e.g., to assess if a group of ground robots provided with the obtained information can pass between obstacles. The map connectivity is analyzed based on a special renumbering of connectivity areas, which is used in data exchange between reconnaissance robots.

This work is focused on the topical problem of developing efficient software tools to design intelligent systems, in particular, integrated expert systems (IESs), which have powerful functionality and scalable architecture. Automated planning methods are employed to facilitate IES development at the design and requirement analysis stages. General problems of developing a modern software tool base for IES design are discussed. In this context, the intelligent software environment of the AT-TECHNOLOGY workbench for IES development based on the problem-oriented methodology is described. The problem of planning the IES prototyping process is investigated, which forms the basis for implementing the basic components of the intelligent software environment (intelligent planner and technological knowledge base). An automation method based on heuristic search is proposed. The use of the intelligent software environment is illustrated by the development of a dynamic IES prototype for controlling and planning medical resources in serious road traffic accidents.