This article centers around the development and analysis of machine learning (ML) and deep learning models aimed at enhancing diabetes diagnosis. In the swiftly evolving landscape of data technologies, it becomes crucial to explore the applications of these methods for accurate predictions and improved medical decision-making. Our research encompasses diverse datasets, leveraging state-of-the-art algorithms and technologies for model training and testing. The primary emphasis lies in evaluating the accuracy, sensitivity, and specificity of models within the realm of diabetes diagnosis. The study results reveal significant advancements in disease prediction, underscoring the potential of ML and deep learning in medical applications. This work introduces fresh perspectives on the utilization of computational methods in healthcare and serves as a foundation for prospective research in this domain.

In this paper, we study the problem of forecasting the energy consumption of real estate objects. The purpose of the work is to increase the efficiency of distribution companies by improving the mechanism for determining planned consumption volumes. The task of the work is to construct a mathematical model based on existing data and further forecast on its basis. To achieve the goal of the work, 2 approaches are used: forecasting energy consumption based on time series (regression task) and forecasting the energy consumption profile for each object (classification task). Various methods of machine learning, as well as clustering, were used in the work. We compared machine learning methods that solve different problems with each other, and selected the best ones. As a result, the ARIMA and Support Vector Machine methods are compared for predicting energy consumption. The choice of a particular algorithm will depend on the goals of the companies that produce and distribute energy.

This article presents a clinical decision support system for diagnosing acid–base balance disorders based on arterial blood gas (ABG) analysis using the Henderson–Hasselbalch equation. Considering the increasing requirements for accuracy and rapid clinical diagnostics, algorithm-based interpretation of ABG parameters has become highly relevant. The study aims to systematize the key blood gas indicators (рН, ???? _???? ???????? ₂ , ???????????? ₃ ⁻ , BE) and analyze their physiological relationships, as well as to develop a logical and mathematical model for automated classification of acid–base disorders. The proposed algorithm enables the identification of metabolic, respiratory, and mixed disturbances, taking compensatory mechanisms into account. The results demonstrate that the implementation of a computerized decision support model enhances diagnostic objectivity, reduces interpretation errors, and improves the quality of clinical decision-making.

Modern technologies have developed quite well lately, and in addition to the banal data collection and processing, modern algorithms are able to predict the behavior of equipment and prevent the occurrence of malfunctions and failures in advance and take actions to reduce the likelihood of an accident and warn the relevant employees. Within the framework of this study, the authors proposed the development of a web-application for the analysis and visualization of data from UK-TPP to ensure the safety of technological processes. Ust-Kamenogorsk TPP is a regional thermal power plant, heat and electricity generation covers 80% of the load of the housing and communal sector of the city and industrial enterprises. The TPP is part of the unified energy system of Kazakhstan. The application being developed has a web interface that will display and visualize data from the main sensors and instruments of the UK TPP, as well as prevent equipment malfunctions and analyze the behavior of sensors using machine learning. Node.js was chosen as the main software platform used, built on the java script programming language, using additional modules that expand the capabilities of this programming language, written in c ++, which speeds up working with data arrays, receiving, and sending them.

The article is devoted to the study of a nitrogen supply system for cooling an experimental device designed for testing in a pulsed reactor with a graphite moderator. To analyze emergency situations caused by disruption of the nitrogen supply system in the cooling paths of the fuel assembly cover and the experimental device power housing, a mathematical model has been developed to describe thermodynamic processes in a system of interconnected tanks that simulate individual sections of the experimental device hydrodynamic cooling system in a pulsed graphite reactor and implemented in the mathematical interactive package for engineering calculations in MATLAB/Simulink. To study the normal operation of the system and analyze possible emergencies, the cooling system of the experimental device was presented in the form of groups of interconnected sections of the hydrodynamic system-tanks, with specified volumes and hydraulic resistances of pipelines connecting these tanks. Verification of the model, calculations of the experimental device cooling system normal operation mode, calculations of emergency situations caused by a violation of the nitrogen supply system normal operation in the cooling paths of the fuel assembly cover were carried out, in particular, it was found that for the cooling system of the experimental device WF-2, the breakdown of the gearbox P-12 can cause depressurization of the fuel assembly cover and an increase in pressure in the power housing, the opening of the safety valve prevents exceeding the pressure limit in the housing, the device maintains integrity even in emergency situations, without being subjected to destruction of the power elements of the structure at a pressure of up to 8 MPa

The aim of this study is to identify the relationship between the level of digital literacy among the population and the innovation activity of enterprises across different regions of Kazakhstan. The research is motivated by the growing significance of digital competencies in enhancing innovation capacity, particularly in the context of regional disparities in technological readiness. A mixed-method approach was applied, including correlation analysis and structural equation modeling (SEM), using regional statistical data for the period 2021-2023. Key variables included R&D expenditures, digital infrastructure accessibility, and indicators of enterprise innovation activity. Data normalization (Z-score) and region grouping were used to ensure the comparability of results across newly established administrative boundaries. The results reveal a nuanced relationship that differs markedly across regions. In major urban centers such as Almaty and Astana, a clear and strong link is observed between digital literacy and innovation activity. In contrast, in many other regions, this relationship appears weak or even negative. Among the most impactful variables are household Internet access, external R&D funding, and the presence of innovation-oriented enterprises. Structural equation modeling highlights the indirect influence of digital infrastructure and innovation density on this relationship. Model fit statistics (CFI = 0.921; p = 0.207) support the adequacy of the analytical framework. These findings point to the importance of developing tailored strategies at the regional level – ones that address specific digital skill gaps and innovation barriers. The study offers practical implications for policymakers and enterprise leaders seeking to strengthen regional innovation ecosystems through targeted digital development.