Authors: Alimbekova N., Hari Mohan Rai, Turymbetov T., Zhumadillayeva A.

Abstract. This study investigates the use of a hybrid neural network architecture that combines convolutional neural networks (CNN) and long short-term memory (LSTM) for efficient sorting of plastic containers. The study focuses on classifying plastic waste based on chemical composition and contamination level, obtained with a near-infrared (NIR) spectroscopy device. Experimental results show that the CNN+LSTM hybrid model achieves relatively high accuracy in recognizing different types and colors of plastics, including the detection of contaminants in containers. A comparative evaluation of the model’s performance was conducted with traditional classification methods such as logistic regression, partial least squares (PLS), and linear discriminant analysis (LDA). The results show that the CNN+LSTM model performs better than traditional approaches, especially in scenarios with small spectral differences between classes. This study demonstrates the potential of machine learning to improve the efficiency of plastic waste sorting and recycling processes, thereby contributing to improved environmental sustainability.

Keywords. Plastic waste, NIRS (Near-infrared spectroscopy), Neural network, Hybrid model, CNN, LSTM.

Authors: Kumarkanova A.,Khasеnova Z.,Weiss Yu.

Abstract. Forecasting drug demand plays a key role in ensuring sustainable supply, effective inventory management, and timely patient access to life-saving medicines. This article presents a study of ARIMA time series and exponential smoothing methods for predicting demand for an antihypertensive blood pressure drug. The research is aimed at developing and identifying a model that ensures high accuracy and efficiency of forecasting based on selected data sets. The analysis includes the study of forecasting methods, data collection and processing, determining the optimal parameters for each method, developing a hybrid model, evaluating accuracy based on specified metrics, and analyzing the results. In the course of the conducted research, it was found that the most effective forecasting methods are time series-based approaches, including ARIMA models and exponential smoothing methods. And the developed hybrid model demonstrates high forecast accuracy by combining the advantages of the two approaches. The results show that the hybrid model is superior to ARIMA and exponential smoothing in all key metrics. Based on the findings, the introduction of hybrid models is proposed to improve the accuracy of forecasting demand in the pharmaceutical industry.

Keywords: forecasting, forecasting methods, exponential smoothing, ARIMA, time series, drug, hybrid model.

Authors: Койшиева Д.Е., Сыдыбаева М.А., Бельгинова С.А., Жаксыбаев А.М., Ерсаинова Ж.Е.

Abstract. Segmentation of medical images is one of the key tasks of biomedical informatics, especially in the context of cancer diagnosis. The task of precise segmentation of cell nuclei is particularly relevant and necessary to identify morphological changes in tissues. However, the main challenges and problems in this area remain the variability of image quality, limited amounts of available data, and the need to ensure the high accuracy and stability of deep learning models. This study analysed the impact of ten data augmentation techniques, including the addition of random noise of varying intensity, on the cell nucleus segmentation model’s performance. The experiment also specialises in techniques using artificial noise addition, simulating real-world conditions such as lighting variations, artefacts, and defects during sample preparation. Two types of noise were used to add: additive Gaussian noise and uniform random noise with an intensity from minus 0.05 to 0.05, which demonstrated a significant effect on the generalizing ability of the model, improving its resistance to heterogeneous data. The architecture used was based on a modified UNet model with the introduction of the CBAM module, which focuses the model’s attention on significant areas of the image. The module has been added to the decoding part of the model architecture. For the experiments, two reference datasets CryoNuSeg and MoNuSeg were combined with preprocessing, including patching and image format conversion. Applied augmentation techniques included horizontal and vertical reflection, random rotation, contrast change, elastic deformation, and noise addition. In addition, augmentation combinations were investigated during the experiments. The study results showed that the techniques of horizontal and vertical reflection augmentation and random rotation achieved minimal losses, while the accuracy of the models exceeded ninety per cent. The contrast change technique demonstrated the most balanced performance, providing an average intersection ratio over the union of 0.928 and an accuracy of 92.2 per cent. The data from the study results emphasize the importance of using artificial noise addition to increase the model’s resistance to artefacts and data variability, as well as the use of combined augmentation techniques. In the future, combined augmentation techniques may become the basis for the development of adaptive algorithms capable of effectively working with heterogeneous biomedical images
Keywords. Histological images, cancer, nuclei segmentation, convolutional neural networks, attention module, augmentation.

Authors: Mekhdiev T.,Shaikhanova A.K., Bekeshova G.B.,Iklasova K.E.,Bakenova K.S.

Abstract. The article discusses methods of intelligent text processing (Text Mining), which allow to transform poorly structured text data into structured and easily analysed information. With the growth of data volumes in the digital age, Text Mining is becoming an indispensable tool for analysing texts in various fields. These technologies find wide application in information security where text analysis helps to identify threats and anomalies, in healthcare to process medical records and extract diagnostic information, in marketing to analyse consumer preferences, and in legal practice where automation of document analysis improves accuracy and reduces time costs. The paper details both traditional statistical methods such as TF-IDF, Word2Vec, Latent Dirichlet Allocation (LDA) and state-of-the-art approaches including deep learning models based on transformer architecture such as BERT, GPT and their derivatives. The state-of-the-art methods show significant advances in context-awareness, semantics analysis, and extraction of hidden meanings from texts, which makes them indispensable for solving complex problems. Particular attention is paid to comparing the effectiveness of different methods and their applicability in automation tasks. The possibilities of Text Mining integration for analysing large amounts of data, identifying patterns and automating knowledge extraction processes are described. The presented research results emphasise the relevance of using these technologies to improve the efficiency of specialists’ work, accelerate the processes of information analysis and problem solving in key industries, which opens new perspectives for the implementation of intelligent data processing systems.

Keywords: Text Mining, intelligent text processing, machine learning, natural language processing, TF-IDF, Word2Vec, BERT, GPT, text analysis automation.

Authors: Nurgaliyeva S.A., Naiman N.B., Adikanova S.S.

Abstract. Plant disease detection is crucial to modern-day agriculture because timely diagnosis can reduce the loss of crops to an appreciable level and improve productivity. This review presents advanced disease detection systems based on machine learning techniques and multimodal data analysis. A comprehensive comparison of different machine learning algorithms, including convolutional neural networks (CNNs), transfer learning models, and object detection methods like YOLO, has been done. This study demonstrates that combining visual data with the analysis of volatile organic compounds (VOC) enhances the accuracy and reliability of the diagnosis. This provides opportunities for the actual development of satellite and cheap systems for monitoring operable in the field. Theoretically, this work contributes to developing strategies for integrating heterogeneous data and optimizing deep neural network models to make them lightweight and effective. The review emphasizes developing scalable and adaptive technologies for plant disease detection within precision agriculture.

Keywords: plant disease detection, machine learning, convolutional neural networks, MobileNet, multimodal data, real-time detection, agricultural technology, VOC sensors.

Authors: Tursunov T.S., Kaibassova D.

Abstract. Recommender systems play a crucial role in personalized content delivery by leveraging user preferences and content attributes. This study evaluates three advanced recommendation models: Neural Collaborative Filtering (NCF), Graph Neural Network-based Content Model (GNN-based Content Model), and Hybrid Neural Network (HNN). Each model integrates deep learning techniques to enhance prediction accuracy and user experience. The NCF model employs a dual-branch structure consisting of Generalized Matrix Factorization (GMF) and a Multi-Layer Perceptron (MLP) to model non-linear user-item interactions. The GNN-based Content Model represents users and items as nodes in a bipartite graph, utilizing Graph Convolutional Networks (GCN) to propagate relational and content-based information across connections. Lastly, the Hybrid Neural Network combines collaborative filtering embeddings with content-based features, aligning content representation within the learned latent space. Our evaluation, based on the MovieLens dataset, demonstrates that the Hybrid Neural Network achieves the highest accuracy (85%), outperforming NCF (80%) and the GNN-based Content Model (77.5%). The hybrid approach benefits from both collaborative and content-driven features, leading to improved user-item match quality. The GNN-based Content Model, despite leveraging structured relationships, struggles with cold-start users due to reliance on content information.
These findings suggest that hybrid approaches are more effective in capturing diverse recommendation signals. Future work may focus on integrating transformer-based architectures and reinforcement learning to further enhance recommendation relevance and adaptability.

Keywords: Recommender systems, Deep learning, Collaborative filtering, Graph neural networks, Hybrid models, Personalization

Authors: Karymsakova I., Bekenova D.,Kozhakhmetova D.,Karmenova M.,Ustinova T.

Absrtact. Today, one of the important problems of the modern world is the increase in the incidence of cancer. Various categories of the population are susceptible to the disease. In this regard, one of the urgent tasks is the early detection of oncological diseases and their prevention. Early detection and accurate diagnosis can increase the likelihood of taking measures to treat and increase the life expectancy of patients.
The use of information technologies in medicine is developing at a high rate. Currently, there are a huge number of decision support systems in medicine in general. Such systems are widely used in the diagnosis of many diseases. Expert systems have also been developed for the diagnosis of oncological diseases.
The main idea of this article is to conduct a scientific analysis and classification of approaches used in expert systems for diagnosis in the treatment of oncological diseases. A special feature is the identification of the most optimal approach for the expert system in making a diagnosis. The problem that this study is aimed at is the justification of the choice of an approach for building an intelligent system that allows to speed up and improve the accuracy of the process of analyzing and making recommendations for the treatment of oncological diseases in the Semipalatinsk region, and the creation of a conceptual model of an intelligent system for making a diagnosis in the treatment of oncological diseases.

Keywords: expert systems, information systems, artificial intelligence, neural networks, analysis, intelligent systems, databases.

Authors: Batalova M.,Paweł Komada,Alibekkyzy K.,Bugubayeva A.,Kunapianova M.

Abstract. The aim of this study is to develop and approve an innovative methodological concept of orthopaedic implant performance monitoring using modern wireless sensor technologies and artificial intelligence algorithms. The paper provides a detailed analysis of the existing methods of diagnostics of implant condition, reveals their technical and methodological limitations, and analyses the current trends in the field of medicine aimed at improving the reliability and safety of implantation. The proposed technique combines the collection of temperature and other physical data by means of high-precision wireless sensors and their processing with the use of machine learning algorithms to predict possible deviations in implant operation and timely detection of the initial signs of wear or damage. The experimental part of the research includes testing of the developed system in real clinical conditions that allowed to obtain significant statistical evidence of its efficiency and accuracy. The results showed that the application of the innovative approach allows not only to improve the accuracy of diagnosis, but also to significantly reduce the response time to potential complications, which is important for surgical intervention and reducing the risk of unfavourable outcomes. The presented approach opens new perspectives for further research in the field of medical implant monitoring, contributing to the development of personalised medicine and improving the quality of life of patients.

Keywords: Orthopedic implants, wireless temperature sensor, artificial intelligence, digitalization, diagnostics, patient.

Authors: Sametova A.,Rashidinov D., Nayzabaeva L.,Suleimenova M.

Abstract. This article considers the process of transition from traditional to electronic document flow in educational institutions. The analysis of functions and processes within the framework of the automated information system of an educational organization is carried out. Key approaches to the implementation of electronic document management from the point of view of the system approach are considered. Based on practical experience, the authors propose the concept of organizing and conducting training in the system of professional development of specialists. Automation of document management in educational institutions is especially relevant in such processes as accounting of incoming and outgoing correspondence, processing of internal documents, coordination of administrative and contractual documents, as well as control of execution of orders. Implementation of electronic document management systems (EDMS) requires integration of directories into a single data repository and synchronization of document processing with other business processes of the institution.

Keywords: office management, electronic document management, management of educational organization, efficiency of public administration, information systems, information technology, electronic management systems.

Authors: Koshekov A.,Toleubekov D., Nautan I.,Kurbanov Y.,Togambayev R.

Abstract. This research explores the creation of a novel training complex that includes an engineering support system specifically designed for the technical management of military and specialized aviation equipment. The importance of this study stems from the need to improve the training efficiency of engineering and technical staff, given the growing technical complexity of aviation equipment and the increasing demands for aviation safety. The main aim of this research is to develop an adaptive educational platform that integrates digital twins of aviation systems, fault diagnosis algorithms, and intelligent methods to tailor the educational process. Within this study, the architecture of a comprehensive engineering support system was devised, which includes data collection, digital modeling, and analytics. An adaptive learning algorithm has been introduced, which takes into account the trainee’s skill level, experience, and progress in mastering the material, thereby automatically adjusting the curriculum. The use of virtual simulators and simulation models enabled the development of a flexible training system that closely mirrors the actual operating conditions of aviation equipment. To evaluate the proposed system’s effectiveness, experimental testing was carried out, comparing the training results of two groups of engineering and technical personnel: the experimental group (using adaptive training with digital models) and the control group (using traditional training methods). The findings showed that the new training complex reduced response time to emergencies by 22%, lowered the number of errors by 30%, and improved the accuracy of procedure execution by 18%. The adaptive training system developed is highly flexible, allowing for customization and integration with modern predictive diagnostics methods for aircraft malfunctions. Implementing such a complex in the training of specialists will enhance aviation safety, reliability, and the economic efficiency of technical aircraft operations.

Keywords: aviation equipment, engineering support, digital twins, adaptive learning, training complex, fault diagnostics.

Authors: Suleimenova M.U., Mukhammedzhanova D.M., Amanova R.T.

Abstract. Premature ageing is one of the most pressing problems of modern medicine. It is associated with accelerated wear and tear of tissues and organs, which can lead to the development of chronic diseases and reduced life expectancy. In recent years, artificial intelligence (AI) has started to play a key role in healthcare, providing new tools for diagnosis and prognosis. This article reviews current AI advances in the study of premature aging processes, including the use of machine learning to analyse aging biomarkers, estimate biological age, and predict the risk of age-related diseases.

Key words: artificial intelligence, premature ageing, biological age, machine learning, biomarkers.

Authors: Mukhammejanova D.M., Suleimenova M.U.

Abstract: The integration of artificial intelligence (AI) into Kazakhstan’s healthcare system presents significant opportunities for enhancing remote patient monitoring, particularly in a country characterized by vast geographic distances and uneven distribution of medical resources. In rural and remote areas with limited access to healthcare, AI solutions have the potential to revolutionize patient care by providing real-time health data. The study aimed to identify the advantages and barriers to using AI for remote patient monitoring in Kazakhstan. Analytical methods were applied in this study, including the analysis of pilot project data focused on the implementation of AI-based monitoring systems. Modern IT solutions were used for the collection, processing, and analysis of medical data, enabling an evaluation of the effectiveness of these technologies. The conducted pilot projects demonstrated a 57% reduction in hospital admissions and a 33% decrease in treatment costs for chronic diseases. The use of AI systems was shown to enable early detection of health issues, reducing the burden on healthcare facilities and improving access to medical services for patients with chronic conditions such as diabetes, hypertension, and cardiovascular diseases. Despite the achieved results, the implementation of AI in Kazakhstan’s healthcare system faces several challenges, including insufficient internet infrastructure in rural areas, data security concerns, and the need for training medical personnel. However, government support and ongoing advancements in AI technologies create opportunities for expanding their application in healthcare. The scientific novelty of the study lies in evaluating the practical effectiveness of AI systems in the context of Kazakhstan, while the significance of the work is reflected in the improvement of medical outcomes and the reduction of healthcare costs.

Keywords: artificial intelligence, remote monitoring, healthcare, telemedicine, Kazakhstan, chronic diseases.