Q.A.Iasaýı atyndaǵy Halyqaralyq qazaq-túrіk ýnıversıtetіnіń habarlary

AUTOMATIC RECOGNITION OF ARTIFICIAL INTELLIGENCE-GENERATED TEXTS. A COMPARATIVE ANALYSIS OF MACHINE LEARNING MODELS

2025-09-16T09:35:34+00:00

. This paper investigates the effectiveness of machine learning methods in automatically distinguishing artificial intelligence (AI)-generated texts from human-written texts. The study was conducted on a balanced dataset (2,750 essays; 1,375 entries per class). 14 linguistic-statistical features were extracted from the text, among which vocabulary_richness, word_count, text_length, sentence_count, and complex_word_ratio were found to have high discriminative value using Cohenʼs d. The features were vectorized using TF-IDF and embeddings, and algorithms such as RandomForest, GradientBoosting, XGBoost, LightGBM, LogisticRegression, SVM, KNN, DecisionTree, AdaBoost, and MLP were evaluated using stratified cross-validation. The results showed that gradient boosting models (especially XGBoost) and transform methods performed well; the classification score on the test set reached very high values. Cluster analysis showed a correlation between thematic structure and class division. However, the generalizability of the obtained high scores requires further testing in the case of cross-domain evaluation, adversarial attacks, and manipulations such as reduction/paraphrasing. Future research is recommended to focus on transformer fine-tuning, adversarial stability, and multilingualism.

IMAGE PROCESSING METHODS USING A PARALLEL ALGORITHM

2025-10-10T03:08:44+00:00

Today, image processing occupies an important place in various scientific and technological fields. The increase in data volume and the need for rapid information processing are pushing for improved processing methods. In this context, the use of parallel algorithms is becoming one of the main ways to increase the efficiency of image processing. This work is devoted to the study of image processing methods that are carried out in parallel. In the course of the work, the basic principles of image processing are considered. In addition, these methods allow you to increase the speed of tasks and use computing resources economically. The basics of parallel processing and its advantages are described in detail. Special attention is paid to parallel computing working with different architectures. Algorithms involving several technologies are especially shown as an example. With a parallel approach, you can apply multiple filters to different parts of the image at the same time. This method improves the overall performance of the system and reduces the time spent on processing. At the same time, the efficient use of devices such as multi-core processors and graphics processors ensures high-quality image processing. Each filter used has its own characteristics and is aimed at specific purposes. Thanks to parallel methods, it becomes possible to quickly and efficiently apply these filters. It plays a big role in real-time video editing. Parallel algorithms with high efficiency significantly improve the result of image processing.

Conditions for solvability of a boundary value problem for ordinary differential equations with nonlocal conditions

2025-08-28T04:25:59+00:00

In this paper, Dzhumabaev's parameterization method is used to study ordinary differential equations with nonlocal boundary conditions. The parameter was introduced and the replacement was performed. The problem under consideration is divided into two parts: the first is the Cauchy problem for an ordinary differential equation, the second is a linear equation with respect to the introduced parameter. To determine the solution to the Cauchy problem, the Mikkusinsky operator method is used, based on the convolution theory. The Mikkusinsky operator method is an effective analytical tool used to solve ordinary differential and integral equations. This method is based on the theory of convolutions and allows finding solutions to equations using operator calculations. The main feature of Mikkusinsky's method is that it is considered algebraically, by transforming the differential equation using specially defined operators. Based on this method, using the values of , we determine the solution to the Cauchy problem and, by placing it on the boundary conditions, we obtain a system of linear equations associated with the introduced parameters . Theorems on the solvability of the original nonlocal boundary value problem are formulated based on the solvability of the resulting equation with respect to its parameters. The results of the study demonstrate effective methods for solving differential equations with nonlocal conditions and clarify their theoretical foundations.

DIRECT PROBLEM FOR THE SPACE-TIME SINGULAR-NONLOCAL DIFFUSION EQUATION

2025-10-15T14:59:47+00:00

This article investigates the direct problem for the space-time singular-nonlocal diffusion equation. The study focuses on solutions of spectral fractional-linear equations associated with the Legendre differential equation. First, the existence and uniqueness of the solution are established by introducing certain regularity conditions. The method of separation of variables is applied to construct the solutions, which are represented as series in terms of orthogonal Legendre polynomials. Furthermore, the convergence of these series is thoroughly analyzed, along with the smoothness and functional properties of the obtained solutions.

The paper discusses Caputo fractional derivatives, Riemann–Liouville integrals, and operators of the involution type. Based on several theorems and lemmas, eigenvalues and eigenfunctions of the spectral problems are derived. The uniform and absolute convergence of the solutions is proven, and their consistency with the given initial-boundary conditions is demonstrated.

Overall, this work contributes to the theory of fractional-order differential equations by proposing new approaches to solving direct problems and laying the foundation for further applications in inverse problem theory. The results can be applied to complex problems in mathematical physics and serve as a basis for use in applied models. Moreover, the proposed methods provide opportunities for improving the numerical solutions of specific problems.

DEVELOPMENT OF A RESEARCH-TRAINING STAND AND INVESTIGATION OF THE RELATIONSHIP BETWEEN THE OUTPUT POWER OF A SOLAR PANEL AND LIGHT INTENSITY

2025-09-29T05:34:05+00:00

Abstract. In the 21st century, the global energy system has been moving towards renewable energy sources. Solar energy, as an environmentally friendly and inexhaustible resource, is becoming a strategic source of electricity production. The efficiency and performance of solar panels are determined by their dependence on external factors, especially the intensity of incident light and the panel’s tilt angle. Therefore, studying the relationship between the output power of a solar panel and light intensity is a relevant scientific direction for optimizing solar energy systems, reducing costs, and increasing efficiency. By developing a special educational and experimental stand, the dependence of the solar panel’s output power on light intensity was experimentally investigated; the panel’s current-voltage (IV) characteristic was plotted, and the fill factor and efficiency indicators were determined. The research was carried out using a research-training stand consisting of a photovoltaic panel, a control panel, projectors, variable loads, voltage/current sensors, and a lux meter. During the experiment, the light intensity was gradually varied, and the panel parameters Voc, Isc, Vmax, and Imax were measured. Based on the obtained data, the fill factor (Fill Factor) and the efficiency coefficient were calculated. The experimental results showed that the output power of the solar panel is approximately directly proportional to the light intensity. The short-circuit current increased linearly with increasing illumination, while the open-circuit voltage changed only slightly. The IV curves at different light levels and the calculated fill factors corresponded to the operating patterns of the photovoltaic panel. In addition, it was observed that the effect of the panel’s tilt angle on power complies with the cosine law. The obtained results can serve as a basis for improving the efficiency of solar panels, ensuring their reliable operation under various conditions, and developing new engineering solutions.

INVESTIGATION OF THE STRUCTURAL PHASE STATE OF HIGH-STRENGTH CR3C2-NICR COATINGS OBTAINED BY HVOF METHOD

2025-09-29T04:10:16+00:00

The study investigated the formation of the composition of a Cr3C2-NiCr coating obtained by the high-velocity oxy-fuel (HVOF) spraying method on the E110 zirconium alloy using structural-phase and microstructural characterization. The aim is to evaluate the technological feasibility of the HVOF process for the zirconium alloy and to identify factors influencing the achievement of the target set of properties. Phase analysis revealed a multiphase state of the coatings, including Cr3C2 carbide, Cr23C6 lower carbide, and CrNi3 intermetallic compound. According to energy-dispersive analysis, a uniform layer with a thickness of 10–78 μm was formed, without columnar morphology and with continuous bonding at the “substrate–coating” interface, indicating satisfactory adhesion. Spectral analysis of sample A showed: Ni – 78.504%, Cr – 17.032%, O – 3.447%, C – 1.017%; sample B showed: Ni – 66.405%, Cr – 29.286%, O – 3.727%, C – 0.581%. The presence of oxygen (3–4%) is associated with spraying in an open atmosphere. In addition, mainly in the transition zone, intergranular and intragranular pores, local microcracks, and layer separation areas were detected; the oxygen content of about 3–4% is due to partial oxidation of the sprayed particles during spraying in an open atmosphere. The overall results confirm the suitability of the HVOF method for obtaining Cr3C2-NiCr coatings on the E110 base and indicate the need to optimize parameters to reduce porosity, limit oxidation, and stabilize the Cr3C2 phase fraction.