Basic methodological and methodical principles used in the creation of the domestic information and analytical platform and corresponding information systems are considered. This methodology is based on application of two novel theories: the theory of proactive (anticipatory) life cycle management of complex technical objects, as well as its complementary theory of multi-criteria evaluation and selection of the most preferred models and polymodel complexes describing the functioning of complex technical objects and corresponding information and analytical systems. The latter theory is called by the authors the qualimetry of models and polymodel complexes. These two theories make a significant contribution to the development of modern computer science: thanks to the theory of proactive life cycle management of complex technical objects, modern computer science is enriched at a constructive level with methods and methdological support developed in the frames of classical cybernetics (generalized control theory); thanks to the developed qualimetry of models and polymodel complexes, a new mathematical apparatus has appeared in computer science, which makes it possible to increase the validity and quality of design solutions in creation of software and mathematical support for information systems, reduce the cost of their design and operation. Brief information about the practical implementation of the developed theories is presented.

Mathematical models are proposed for estimating the contribution of the cost of the operating system to the total cost of ownership of geographically distributed technical complexes. The basic strategies in organization of the complexes maintenance are considered — planned calendar maintenance and maintenance for fixing failures. The results obtained can be used for preliminary calculation of operational costs and clarification of the cost of the life cycle of complexes during their development or acquisition.

The problem of combining models of classification and ordering of objects within one decision support system is considered. The problem is solved by searching for similarities and differences of the corresponding models using the system analysis methods. Relations between classification models with clear and fuzzy class boundaries, models of ordering objects with hard and soft restrictions on the indicators’ values are considered. The concept of a real goal is introduced, which makes it possible to link the methods of multi-objective and conditional optimization, and these, in turn, with the methods of the multidimensional theory of value and utility. Interpreted as a boundary between classes, the real goal allows to link the methods of classifying and emphasizing objects. The revealed connections are the basis for combining the considered methods within the framework of a common decision support system, which gives great opportunities for making decisions using different methods and comparing the results obtained.

An approach to planning the reconfiguration of complex multi-mode objects under conditions of a given or unknown cyclogram of operating modes is considered. The proposed approach is based on the concept of complex (system) modeling, concept of parametric genome structure, taking into account the functional and technological features of the object under study. Along with the classical "blind" reconfiguration technology, the technology of structural and functional reconfiguration is considered, which implements the change of the structural states of the studied objects both in the event of emergency situations and in the process of planned functioning in order to improve the reliability and increase the duration of the functioning of objects due to the uniform load of elements. A comparative analysis of the results of planning the reconfiguration of complex multimode objects based on these technologies is carried out, advantages of the technology of structural and functional reconfiguration are demonstrated.

An example of creating a subsystem of managerial decision-making in the general structure of maintaining vertically integrated objects of economic activity in a working state as an element of a complex and purposefully complicated architecture of information and control systems is considered. The proposed approach is based on the complexity management methodology developed by the authors, which justifies the order of systematic and evolutionary replacement of decision makers at production facilities with functionally equivalent information-control software and hardware complexes. One of the most effective methods of forming a transparent mechanism for managerial decision-making that can become the main "decisive core" (knowledge base) for such complexes is the "fuzzy-possibilistic approach". The effectiveness of this approach lies in the possibility of using expert information to restore a complex multi-connected functional relationship between the actual production and technological characteristics of the process under study and management decisions made on their basis. An example of creating a knowledge base of control model for a process with a "fluidized bed" is given.

Harvesting of grass forage is considered as a complex technical and technological process. The final product, in this case a silage billet, can be obtained using different variants of feed production technologies. An approach to choosing the most acceptable variant of technologies and plans for the production of grass feed based on logical-dynamic and fuzzy-probability models is presented. The developed set of models and algorithms makes it possible to describe and investigate the existing interconnected multi-stage processes of grass feed production from a system-cybernetic standpoint when transferring a management object (an agricultural enterprise producing feed) from a given initial state to a given final state, depending on specific scenarios of changing external conditions. Factors related to agrobiological, temporal, climatic, economic and organizational constraints are considered as the main external conditions. To solve the problems, fuzzy-probability models for estimating the yield of forage lands and the quality of the grown fodder mass are proposed. When analyzing the effectiveness of feed production processes, an important issue is the multi-criteria assessment of quality, as well as the corresponding programs for proactive management of the silage harvesting process, taking into account the following indicators: the timing of operations under various conditions, the degree of uniformity of resource use and the total time of harvesting grass feed. An original mathematical analysis of the processes under consideration and algorithms for solving problems of forecasting and planning forage harvesting are proposed.

The features of development of open-type Internet monitoring systems with an unlimited number of sources in conditions of a limited amount of data storage systems are considered. The purpose of the work is to solve the problem of forming a set of documents of the minimum required size (the core of documents) that meets the requirements of representativeness and variability of topics when monitoring the Internet. To formalize and solve the problem, a set-theoretic model of the document core is developed. The proposed approach is distinguished by the use of a preemptive algorithm that supports the availability of only relevant documents in the database within the available volume of the data storage system. The results of an experiment using real data confirming the applicability of the developed model are presented. The proposed approach can be used in a number of practical tasks, in particular for searching the Internet for information (documents, pages) for which there is no a priori information needed for keyword search.

To solve the problem of information security event correlation, a model for the combined use of intelligent correlation methods is proposed. Intelligent security event correlation methods are able to analyze both historical data and real-time events and automatically detect changing thresholds. The proposed model contains two levels of data processing: the level of knowledge representation and the level of security event correlation. At the level of knowledge representation, structural and semantic analysis of events is carried out. At the correlation level, the similarity assessment of elements of security event vectors, a graph-oriented neural network method and data analysis using recurrent neural networks are used for event processing. The results of the model are the sequence of interrelated security events, the type of the current security state of the system and the predicted states. The performance of the approach based on the proposed model is illustrated by results of an experiment on predicting system security events, showing low values of the error indicator.

A new approach to solving the problem of automatic face recognition of people using personal protective equipment such as a medical mask has been proposed and tested. This approach is based on the use of methods of generating synthetic images of partially hidden faces and the face recognition model ArcFace. A strategy for training data sets formation is proposed and a number of corresponding recognition models are derived. A series of experiments aimed at assessing the quality of predictions of the obtained solution are carried out, and a relationship between the resulting quality of predictions implemented by recognition models and the volume of synthetic images in training datasets is established. According to the results of experimental studies, neural network models, further trained on datasets with volume of artificially synthesized images of 40-60%, demonstrate values of recognition accuracy above 87% on the AAc quantitative metric (Average Accuracy). Using the proposed approach makes it possible to significantly improve the quality of recognition of partially hidden faces compared to the basic approach.