The paper considers the use of the well-known House of Quality (HoQ) model as a template for structural modeling of cascade failures occurring in interacting networks. The use of HoQ makes it possible to model not only sudden, but also gradual failures in a network system. A distinctive feature of the HoQ model is that its use allows not only to formalize the procedure for constructing a cascade of failures, but also to single out dynamically emerging and disappearing contours, which may include components of both the same and different networks.

The article discusses the operation of robotic manipulation systems for the most popular tasks in the industry. The article provides an implementation of classical methods for grasping objects using a CAD model of the object, highlights their advantages and disadvantages. As a new solution, it is proposed to use a system based on the Action Chunking with Transformers (ACT) neural network architecture. The article details the use of ACT neural networks, the algorithm for training neural networks and launching them within the framework of technological operations of real production. The paper describes the hardware of the system, which includes the ARM95 Collaborative Manipulator, the RealSense Depth Camera D405 depth camera and the HTC VIVE position tracker. The following technological operations were considered as an experimental part of the work: grasping a box object, grasping a pencil object, painting a part and grinding a surface. The developed system shows that modern technologies, including machine learning methods, help to solve complex technological operations with a high level of productivity.

The article investigates the application of ensemble machine learning methods for diagnosing and forecasting the states of technical and technological objects under conditions of noisy data, nonlinear dependencies, and high-dimensional feature spaces. The relevance of the work is driven by the need to enhance the reliability of industrial systems through minimizing accident risks and optimizing operational processes. Traditional approaches demonstrate insufficient accuracy in complex scenarios, motivating the use of ensemble technologies that combine predictions from multiple models to achieve robust results. The primary focus is on Bagging, Boosting, and Stacking methods, their mathematical foundations, and practical implementation. An experiment was conducted using an ensemble of convolutional neural networks (CNNs) for classifying defects in metal microstructure. The results showed an increase in prediction accuracy with an increasing number of classes compared to a single model, confirming the effectiveness of ensembles in reducing error variance and correcting model bias. The proposed approach demonstrates potential for integration into industrial systems, enhancing diagnostic reliability and operational safety of complex technical systems.

This paper addresses the problem of estimating parameters of stochastic event flows based on sample data using machine learning methods. Event flows, characterized by random intervals between the moments of occurrence, are widely used in the modeling of network traffic, telecommunications, computing systems, and in queuing theory. Accurate estimation of such flow parameters is crucial for subsequent analysis, forecasting, and load management in systems with uncertain input information. As training data for the models, we used event arrival times from two types of streams: a Poisson flow (with inter-arrival times following the exponential distribution) and a renewal process (with inter-arrival times following one of twelve probability distributions: gamma, hyperexponential, lognormal, uniform, inverse gamma, Weibull, Pareto, Lévy, Fisher, Fréchet, Lomax, and Burr XII). These distributions were selected due to their diverse statistical properties (presence or absence of moments, asymmetry, heavy tails), which enables coverage of a broad range of applicable scenarios. To solve the parameter estimation task, we employed fully connected neural networks and the CatBoost implementation of the gradient boosting algorithm. As input features for the models, we used the inter-arrival times and their numerical characteristics: mean, standard deviation, variance, coefficient of variation, and quantiles of various levels. To evaluate the model performance, classical machine learning metrics were used: MAE, RMSE, and R². The study also included an assessment of the importance of features used in training. This was done using built-in interpretation tools of gradient boosting, which allow for a quantitative analysis of each feature's contribution to the parameter estimation.

The global surge in the prevalence of depression, characterized by persistent sadness, disinterest, and decreased functioning, highlights the shortcomings of prevailing diagnostic and treatment paradigms. This underscores the urgent need for enhanced interventions, given the inherent limitations of traditional approaches to diagnosing depression. Recent advances in artificial intelligence applications have sparked growing interest in the development of automated depression diagnostic systems among emotion computing experts. The emergence of large-scale language models, such as BERT and its derivatives, for text-based depression detection demonstrates the need for multimodal approaches that integrate text and audio data to achieve more accurate diagnosis. Here, we explore the capabilities of existing large-scale language models and present a proposed multi-stream model, the Multi-Stream Mood Insight Encoder (MMIE). MMIE is designed to seamlessly utilize integrated text and audio data streams with processing capabilities via the Reformer encoder. As part of this concept, linguistic features such as absolutist words and first-person pronouns were incorporated into the Reformer encoder. This holistic approach facilitated a comprehensive analysis of a person's mood and emotional state. Experiments demonstrated that the ClinicalBERT language model outperformed the proposed binary depression classification model. Subsequently, the sigmoid values of the Reformer model were used to diagnose depression. Using the proposed model, experiments were conducted on the DAIC-WOZ dataset. The results showed significant improvements, demonstrating an F1 of 0.9538 for classification, an MAE of 3.42, and an RMSE of 4.64 for regression compared to state-of-the-art methods. These results demonstrate the effectiveness of the proposed model in facilitating the diagnosis of depression.

The paper proposes a new mathematical model of a combinatorial type of product. It is formed on the basis of all possible correct cuts of a hypergraph of a mechanical structure into two connected subgraphs and is represented as an AND–OR-tree. This tree describes all the connected and coordinated product fragments and the inclusion ratio of such fragments. The model can be used to synthesize various design solutions for the technical preparation of discrete production: assembly and disassembly plans, breakdowns into assembly units, etc. AND – OR-the tree of cuts provides objective information for in-depth structural analysis of complex technical systems.

The paper proposes a study of a mathematical model describing the operation of virtual machines in a cloud node. The mathematical model is presented as a queueing system with an infinite number of servers, each of which corresponds to a single virtual machine in a node. The lifecycle start times of each machine in a node are modeled by a Markovian Modulated Poisson Process, which takes into account the correlated nature of their launches. A distinctive feature of this model is the dependence of the operational intensity of each virtual machine on their total number in the node. This effect is called "service rate degradation," and it allows the model to account for competition for resources in cloud nodes. While this allows for a more accurate description of real computing  systems, the dependence of service rate on the number of virtual machines in a system significantly complicates the study of such mathematical models. This leads to the need to develop new methods for analyzing systems taking into account the degradation of the service rate and correlations between incoming requests. The purpose of the proposed work is to obtain analytical expressions for calculating the probabilistic characteristics of a cloud node in the form of a queueing system with the rate of processing requests depending on their total number in the system.

Ensuring a high level of information security for the tax authorities of the Russian Federation is an urgent task. Leaks of personal data from tax service systems can cause serious harm to both individuals and legal entities. Quite often, such leaks occur due to the fault of employees. So it is important to constantly carry out work aimed at improving the computer literacy of tax authorities’ users in the field of handling protected information. The purpose of this work is to construct a logistic regression model for the competence of specialists in information security issues at one tax service in the Irkutsk region. When modeling to identify unknown parameters in logistic regressions, maximum likelihood method and method based on solving linear programming problem were used. To verify the models, a classification accuracy criterion was used as well as ROC analysis elements in the form of curves reflecting a compromise between sensitivity and specificity at different classification thresholds. The interpretation of regression estimates was carried out using the odds ratio. Specialized tests were developed to assess the degree of competence of employees in the tax service. The results of model identification showed that the method based on linear programming had better classification accuracy in all cases compared to the maximum likelihood method. The best logistic regression revealed the factors that have the greatest impact on the level of competence among tax service employees. This made it possible to develop training materials for specific categories of employees. As a result of the training events, the number of employees who successfully overcame the 70% threshold increased from 90 to 150 out of a total of 160 people.

An information system is proposed that allows for the prompt presentation of numerical results of ENB indicators with the selection of the optimal option for building a decentralized energy complex. The system's architecture enables the simulation of various situations and the visualization of indicative indicators of ENB for a territory or facility, such as the consequences of emergencies, changes in the structure of autonomous power supply systems, including the introduction of renewable energy sources, etc. The combination of selected modules in the developed system for monitoring the energy security of remote and isolated territories allows for the systematic monitoring of its level. The final stage of the information and analytical system's functioning is the determination of rational and effective measures for strengthening energy security. The studied energy zones, with their specific characteristics, do not have the necessary ability to "resist" the effects of natural or territorial energy security threats. Therefore, the system's alarm module for the level of the ENB state allows us to assess the degree of the crisis situation and suggest a transition to one of the types of mechanisms for improving it. Operational monitoring is accompanied by the determination of the expected individual results for the proposed set of measures. The developed information and analytical system allows us to assess the level of the ENB in an integrated environment and calculate the current indicators of the ENB, taking into account the specifics of the decentralized territories of the North and the Arctic zones. The algorithm for identifying and assessing the weak and vulnerable points of the ENB ensures the identification of priority areas with strict requirements for ensuring comfortable living conditions for electricity consumers. The model and diagnostic tools for decentralized areas can be used both on existing decentralized facilities and territories, and at the pre-investment stage of energy security improvement measures.

The relevance of studying threats to energy security in modern conditions is increasing and imposes new requirements on models and tools for conducting research on analyzing the impact of threats to the normal functioning and development of energy industries and ensuring reliable energy supply to consumers. Taking into account external and internal factors forms the requirements for the applied models and tools for conducting research on assessing the impact of threats on the energy supplies reliability to consumers. The article presents a description of the factors taken into account in the model, a diagram of working with the model and a test example.

Relevance. The joint work of intelligent agents within an organization makes it relevant to research on the development of methods for coordinating the perception of a situation of purposeful state, goals, interests, values, norms and the creation of common ideas about the situation of choice. The purpose of the work. Using the principle of parallel perception, we propose an approach and a model for coordinating perception and ideas about the choice situation. This ensures: a) expanding the range of approaches in shaping strategies and tactics for the evolution of the organizational system; b) early detection and understanding of trends and the formation of a coherent vision of the future; d) a basis for experimentation and anticipation of new ideas; e) building capacity for self-development; f) strengthening and developing the intellectual capital of the organization; h) a coherent view of the problems. The methods of the theory of fuzzy sets, the theory of active systems, the theory of fuzzy models and networks are used. The main results. The concept of a subjective view of a subject area is introduced. It is proposed to use a fuzzy product model to formalize the ideas. The paper describes the structure of representations and one of the options for its adaptation. The methods of evaluating the usefulness of the representation model and its formalization are considered. An approach based on the hypothesis of common goals and interests of the personnel (agents) of the system is proposed to describe the organization's learning model. The validity of this hypothesis follows from the obvious fact that in an organization, due to the division of labor, the result of each depends on the joint efforts of all. Consequently, the individual contributions of agents based on insight, intuition, knowledge, experience, and available information of a particular agent with differing values, norms, beliefs, goal structures, and abilities, when they come together to do some work, determine the future outcome.

This study examines social dilemmas arising from the introduction of autonomous vehicles (AVs) into a heterogeneous traffic flow that includes human-driven personal vehicles and buses. The analysis employs a traffic flow model based on the Revised S-NFS cellular automaton, which captures the interaction effects among different types of agents. The presence of a social dilemma is identified through the Social Efficiency Deficit, defined as the gap between the Nash equilibrium and the Pareto-optimal distribution of strategies. Several scenarios with varying AV penetration rates and initial traffic densities are considered. The results demonstrate that both the prisoner’s dilemma and the hawk–dove game may emerge between buses and other vehicle types, while AVs consistently exhibit a stable speed advantage, leading to the disappearance of social dilemmas once they dominate the flow. These findings enhance the understanding of how AVs affect transport mode choice dynamics and provide a basis for shaping transport policies that account for the conflict between individual and collective interests.

The article presents the experience of creating digital twins of cultural heritage objects of the Krasnodar region, geoinformational analysis of the obtained models. The main purpose of this article is to assess the prospects of development of the concept of digital doubles of cultural heritage objects based on the study of photogrammetric technology of their creation on the example of cultural heritage objects of Krasnodar Krai. Possible ways of use and prospects of development of the created digital doubles are defined.

The computer-aided design (CAD) system used at an enterprise to create a digital model of structures or products is a set of several software products, each of which defines strictly defined functions as part of the enterprise's production system.
A rational choice of CAD begins at the stage of preparation for creating a digital model. The choice should take into account: the software (SW) required for 3D modeling of the finished object, as well as the final software intended for printing the entire set of design documentation, or for generating an electronic set of documentation for sending to the customer.
In the context of sanctions from foreign countries, it is necessary to approach the selection of software products, taking into account all their features, advantages and functionality. In addition, it is necessary to consider the possibility of organizing technical support for these software products, as well as their adaptation to the domestic software that is used to compile auxiliary documents included in the design documentation set.
For the effective organization of work between individual software in order to enable their interaction with each other, it is necessary to develop a methodology that would reduce the interaction time and implement interaction mechanisms. These mechanisms provide for the transfer of information from one software package to another for the purpose of processing information and automatic generation of additional documents (estimates, statements of work volumes, specifications, etc.).
In this regard, the development of a new method for selecting CAD and the formation of basic principles for information transfer is very relevant.

The issues of mental and psychological health of adults, children and adolescents are becoming particularly relevant in the modern world. Urbanization, acceleration of the pace of life, digitalization of communication, mediatization of the educational environment and increasing demands on the volume of information processed by students of modern educational organizations can affect their emotional state, which can lead to various emotional disorders, the absence of which is one of the criteria for mental health. In the paper we present the results of the developing and programming of a chatbot for diagnosing the parameters of the respondent's mental health. The neural network was trained to recognize signs of depressive, anxious or aggressive tendencies of the respondent in the dialogue. In each case, neural network offers its own version of diagnostic instruments and recommendations for a respondent. We developed telegram chatbot to identify signs of depression, auto-aggression, or anxiety in the respondent's statements during communication. When identifying signs of risk, the chatbot offers a number of questions and, with informed consent, to take a psychological test. Several psychological tests were uploaded: the A. Beck Depression Scale; the A. Beck Anxiety Scale; the Auto– and Heteroaggression Questionnaire by E. P. Ilyin. The GigaChat neural network from the Sber company was used to communicate with the user. The programs were written in Python. This program can be used by educational organizations for monitoring the parameters of the emotional state of students.