Abstract : This article provides an overview of modern reinforcement learning (RL) algorithms used for solving real-time tasks. Various methods, including Q-learning, gradient algorithms, recurrent neural networks, and distributed learning, are analyzed, highlighting their capacity to adapt to changing environments and make effective decisions. Special attention is paid to computational resource optimization and model stability, which are critical for tasks requiring rapid response. Knowledge adaptation and transfer methods, such as multi-task learning, are also discussed as approaches that accelerate model training in data-scarce conditions. The application of these methods makes RL an effective tool for dynamic, high-tech fields, including autonomous control and robotics. The findings demonstrate the potential of RL in real-time conditions but also underline the need for further research to enhance algorithm resilience and reduce computational costs.

Abstract : This article presents an overview of synthetic data generation methods for training artificial neural networks (ANNs) under limited access to real data. Key approaches such as Generative Adversarial Networks (GAN), data augmentation methods, and statistical models applicable to various data types are reviewed. Special attention is given to the application of synthetic data in face recognition, rare disease diagnosis, and autonomous systems management. The advantages and limitations of each method, as well as their impact on model accuracy, are analyzed. Potential risks associated with synthetic data, including biases and distortions, and approaches to mitigate these issues to enhance model reliability, are also discussed. The use of synthetic data provides substantial opportunities for advancing ANNs, improving their effectiveness and generalizability in practical tasks.

Abstract : This article presents a comparative analysis of blockchain platforms used for financial transactions, aimed at identifying their key characteristics and application areas. Platforms such as Ethereum, Hyperledger Fabric, Ripple, and Stellar are reviewed, with differences in consensus algorithms, transaction processing times, throughput, and decentralization levels highlighted. Security and resilience aspects affecting the effectiveness of blockchain solutions in the financial sector are discussed, along with prospects for using various platforms in both open and corporate systems. Special focus is placed on analyzing scalability, energy consumption, and integration with existing financial services. The study results indicate that the choice of blockchain platform depends on task specifics and requirements for security, speed, and decentralization level. Blockchain applications in the financial sector are evolving, providing opportunities to enhance the reliability and efficiency of transactions.

Abstract : This article provides an analysis of the potential of decentralized networks (DN) for supply chain management. Key benefits of this technology, including transparency, reliability, and reduced transaction costs, are discussed. Examples of successful DN implementation by large companies such as Maersk and Walmart are presented, illustrating the impact of this technology on logistics efficiency. DN enables real-time data sharing between participants, improving coordination and inventory management. The article also addresses limitations and challenges, such as low data processing speeds and integration difficulties with existing systems. It is anticipated that further development of decentralized technologies will promote their broader adoption in logistics processes, enhancing supply chain transparency and security.

Abstract : This article explores the impact of containerization on IT project infrastructure, analyzing its benefits and limitations. Containerization simplifies dependency management, enhances flexibility, and speeds up deployment by isolating environments. Examples from major companies such as Netflix, Spotify, and Google highlight its advantages in scaling and adapting infrastructure to high loads. Special attention is given to container orchestration using Kubernetes, which facilitates managing microservices architecture. The limitations of containerization, including the need for improved security and data management, are also discussed. It is anticipated that containerization will become an integral part of IT infrastructure, especially within DevOps environments.

Abstract : The article examines key technologies for storing and processing unstructured data within Big Data environments, including Hadoop, NoSQL databases, Apache Spark, and Elasticsearch. Key advantages and limitations of each approach, along with their impact on infrastructure performance and scalability, are analyzed. An example is provided using Apache Kafka for data streaming and PySpark for preprocessing, highlighting the significance of these technologies in handling large volumes of information. Recommendations are given for selecting suitable technologies for different business scenarios. The research demonstrates that the integration of Big Data technologies into business processes enhances flexibility and reduces costs associated with data storage and processing.

Abstract : Intelligent monitoring systems play a pivotal role in the development of smart cities by providing data collection and analysis necessary for efficient urban infrastructure management. The primary goal of this article is to examine modern approaches to designing intelligent monitoring systems and to assess their potential for urban use. Issues related to real-time data integration and processing are explored, as well as the application of machine learning and big data technologies for process automation. The article also discusses major challenges in implementing these systems, such as data security, infrastructure costs, and legal compliance. The reviewed approaches and methods highlight key factors for successful deployment of intelligent monitoring systems in cities aiming for sustainable development.

Abstract : This paper examines the main security protocols used in the Internet of Things (IoT) and analyzes methods aimed at improving data and device protection. With the growth of IoT devices, the risk of cyberattacks increases, necessitating specialized protocols like TLS, DTLS, and MQTT. TLS and DTLS provide reliable data encryption at the transport layer, though their high resource requirements limit usage in low-power devices. MQTT, optimized for low-resource devices, supports built-in authentication and encryption functions, making it popular for IoT networks. The paper also considers lightweight cryptography to enhance security with limited computational capacity and distributed access management systems based on blockchain. The combination of security protocols and adaptive methods achieves high resilience in IoT networks, enhancing overall system security and reliability.