This thesis covers the theoretical and practical aspects of improving the methodology for increasing the physical fitness of female handball players. Within the framework of the study, an approach was developed that focuses on the differential distribution of training loads, the integration of technical and tactical components in the development of physical qualities, and the balancing of load-recovery processes, taking into account the physiological characteristics of the female body. Experimental observations were conducted in three regions, and significant positive changes were noted in the experimental groups compared to the control groups (p<0.05). The results obtained confirm the effectiveness of individualized methodological approaches in women's sports.

Stroke remains one of the leading causes of mortality and long-term disability worldwide, creating substantial clinical, economic, and social burdens. Despite significant advances in prevention and treatment, the increasing prevalence of cardiovascular diseases, aging populations, and lifestyle-related risk factors continue to elevate stroke incidence across diverse populations. Traditional stroke risk assessment approaches rely primarily on statistical models and clinician-based evaluations; however, these methods often struggle to capture complex nonlinear interactions among multiple risk determinants. The emergence of machine learning (ML) has introduced innovative opportunities for improving predictive accuracy, enabling personalized risk assessment, and supporting proactive healthcare interventions. This systematic review critically examines the role of machine learning–driven frameworks in predicting stroke risk factors, synthesizing evidence from contemporary studies addressing epidemiological trends, clinical determinants, predictive analytics, and computational modeling approaches. The review analyzes existing machine learning methodologies, including nonlinear predictive systems, cardiovascular risk-based frameworks, and healthcare analytics models applied to stroke prediction. Furthermore, it evaluates the strengths, limitations, and practical implications of ML-based risk prediction systems within clinical environments. Findings indicate that machine learning models consistently outperform conventional risk stratification approaches in handling multidimensional datasets, identifying hidden relationships among variables, and supporting individualized prediction. However, challenges related to data quality, model interpretability, generalizability, and clinical integration remain significant barriers to widespread adoption. The review proposes a conceptual framework integrating epidemiological, clinical, and computational dimensions to strengthen future stroke prediction systems. The study contributes to the growing literature on intelligent healthcare analytics by providing a comprehensive assessment of machine learning applications for stroke risk prediction and identifying future directions for research and implementation.

A mathematical model of the statics of the preliminary distillation of cottonseed oil miscella in tubular apparatuses with an ascending liquid film is proposed. It facilitates the design of apparatuses with an optimal heating surface and the identification of their effective operating modes based on an analysis of the temperature and miscella concentration distribution along the height of the tube bundle. Selected results of a study of the statics of cottonseed oil miscella evaporation in a tubular evaporator of the Extraction Technician extraction unit of Gulistan Extract-Yog JSC are presented. It was found that, at a miscella flow rate of Gо = 3,31 t/h, to achieve the required outlet concentration аex ≥ 95 % at temperatures tex ≤ 110-115 ℃, 76÷84 m2 of the active heat-transfer surface of the apparatus is required. To increase the efficiency of the device, it is recommended to increase the miscella consumption by 13÷24 % in accordance with the identified heating surface reserve (24 m2).

The increasing rate and intensity of ransomware attacks have become an important economic risk to contemporary businesses, and to mitigate it, strong analytical frameworks are required to measure financial loss and inform strategic investment in cyber resilience. This paper builds a predictive analytics-based modeling framework to provide an approximation of the economic impact of ransomware events and optimize business continuity planning. The study uses frequency severity modeling and machine learning to estimate direct costs (e.g., payments to ransoms, data restoration, etc.) and indirect losses (e.g., time out of business, branding loss, etc.) based on aggregated industry-related datasets in the form of the industry reports, such as IBM Cost of a Data Breach, Verizon Data breach investigations Report, Sophos State of Ransomware. The study methodologically uses a hybrid methodology that combines probabilistic risk modeling and supervised learning algorithms to predict the expected loss distributions in different scenarios of threat. Simulations involving scenarios are used to determine the return on investment (ROI) of cybersecurity controls such as backup systems, employee awareness training, and zero-trust architectures. The results indicate that proactive resilience strategies in organizations can mitigate projected ransomware-related losses by up to 40%-60%, with diminishing marginal returns of the level of security investment. The research adds value to the literature by filling the gap between technical cybersecurity analytics and economic decision-making by providing a scalable model to quantify enterprise risk. In practice, it offers practical decision-makers with insights on the best resource allocation to invest in cybersecurity and boost organizational preparedness to the changing ransomware threats. The fact that the model can be applied to a range of industries also supports its applicability to policymakers and corporate leaders who may want to enhance digital resilience in a more hostile cyber environment.

Modern enterprises are increasingly dependent on scalable data analytics systems capable of handling large, heterogeneous, and rapidly changing datasets. While cloud-based big data platforms dominate this landscape, spreadsheet environments—particularly Microsoft Excel enhanced with Power Pivot—continue to play a critical role in enterprise reporting due to their accessibility, flexibility, and widespread adoption. This paper examines the scalability of spreadsheet-based analytics with a focus on Microsoft Excel Power Pivot as a semantic modeling and in-memory analytics engine. It evaluates how Power Pivot extends traditional spreadsheet limitations by integrating relational data modeling, in-memory compression, and advanced calculations using DAX. The study situates Power Pivot within the broader ecosystem of big data analytics, comparing it conceptually with distributed frameworks such as MapReduce and Spark-based systems. It also explores governance, data integration, and time-series analytics challenges in enterprise environments. Drawing from established literature in business intelligence, data governance, machine learning, and distributed computing, the paper argues that Excel Power Pivot remains a relevant and scalable solution for mid-tier enterprise analytics when properly architected. The discussion highlights trade-offs between usability and scalability, and proposes hybrid architectures combining spreadsheet models with enterprise-grade data pipelines.

This study presents a multi-index remote sensing assessment of desertification dynamics in the Khorezm irrigated zone of the Amu Darya Basin (59–62.5°E, 40.5–42.5°N) over the period 2000–2018, using Landsat Collection 2 Surface Reflectance imagery. Four spectral indices — the Bare Soil Index (BSI), Soil-Adjusted Vegetation Index (SAVI), Normalized Difference Water Index (NDWI), and Normalized Difference Drought Index (NDDI) — were computed annually and interpreted alongside NDVI-based Vegetation Condition Index (VCI) drought classification. BSI-based desertification status classification reveals that Critical desertification conditions (BSI > 0.20) persisted across the majority of study years, including years simultaneously classified as No drought by VCI, demonstrating that vegetation greenness recovery does not eliminate underlying soil degradation. SAVI analysis confirms that standard NDVI underestimates vegetation stress in the semi-arid Khorezm landscape by approximately 30% during drought years due to soil background reflectance contamination, with the SAVI-to-NDVI best-to-worst year ratio reaching 2.5:1 versus 1.9:1 respectively. NDWI values remain consistently negative across all years, with the most negative values paradoxically occurring in the best vegetation condition years, reflecting the spectral dominance of dense crop canopy NIR reflectance. These findings demonstrate the necessity of multi-index monitoring approaches for accurate desertification assessment in irrigated dryland systems and provide direct methodological support for the implementation of Uzbekistan's Land Degradation Road Map for 2024–2028.

The modernization of enterprise banking systems has become a strategic requirement as financial institutions increasingly depend on high-availability, low-latency, and secure digital services. Traditional monolithic banking applications often struggle to support real-time transaction processing, elastic scalability, fault isolation, and rapid feature delivery. This research paper proposes a cloud-native, event-driven enterprise banking platform designed using Java microservices, Kafka-based asynchronous communication, secure API gateways, automated DevOps pipelines, and infrastructure-as-code practices. The study develops a conceptual and technical architecture for transforming legacy banking systems into distributed real-time ecosystems capable of handling payments, account services, fraud monitoring, notification workflows, audit logging, and regulatory reporting.

The proposed model is grounded in prior research on serverless and cloud-native computing, function composition, platform performance, fraud detection, anomaly identification, and distributed enterprise security. Existing literature highlights the advantages of cloud programming abstraction, scalable service deployment, and event-based processing, while also revealing unresolved challenges related to latency, observability, false positives in financial monitoring, and operational complexity (Jonas et al., 2019; Baldini et al., 2017; Castro et al., 2019; Wang et al., 2018). In response, this paper presents an integrated architecture that combines domain-driven microservice decomposition, Kafka event streams, Java-based service orchestration, CI/CD automation, containerized deployment, and real-time analytics pipelines.

The findings indicate that event-driven microservices can improve transaction responsiveness, reduce dependency coupling, and enhance system resilience when compared with tightly integrated legacy architectures. Kafka enables durable, scalable, and replayable event processing, while automated deployment pipelines reduce release friction and support continuous modernization. However, the model also introduces governance challenges involving event schema management, distributed tracing, data consistency, security enforcement, and operational monitoring. The study concludes that cloud-native event-driven banking platforms offer significant potential for mission-critical financial modernization, provided that performance engineering, fraud controls, DevSecOps, and compliance-by-design principles are embedded into the platform architecture.

Psychological readiness is considered one of the key determinants of successful athletic performance, particularly in technically complex track and field events such as the hammer throw. The ability of athletes to effectively manage competitive stress, maintain concentration, regulate emotions, and sustain self-confidence significantly influences the quality of technical execution and competitive outcomes. Despite the growing attention to psychological preparation in modern sports science, limited research has specifically examined its role in competitive hammer throw performance among university athletes.

The purpose of this study was to investigate the influence of psychological readiness on the competitive performance of hammer throw athletes and to identify the psychological factors that contribute most significantly to successful performance during competitions. The research employed a combination of theoretical analysis, pedagogical observation, psychological assessment, and statistical methods. Psychological indicators such as competitive anxiety, self-confidence, emotional stability, concentration, and motivation were evaluated and analyzed in relation to athletes’ competitive results.

The findings indicated that athletes with higher levels of psychological readiness demonstrated superior technical consistency, better emotional control, and improved competitive performance. In particular, self-confidence and concentration were identified as the strongest predictors of successful hammer throw results, while elevated levels of competitive anxiety were associated with decreased technical efficiency and performance outcomes. Furthermore, psychological readiness was found to facilitate more effective execution of technical skills under competitive pressure.

The study confirms the importance of integrating psychological preparation into the training process of hammer throw athletes. The results provide theoretical and practical evidence supporting the development of comprehensive psychological training programs aimed at enhancing competitive performance and overall athletic achievement.

Human echinococcosis, a dangerous zoonotic parasitic disease that is still a major public health concern in many developing countries, is caused by the larval stages of Echinococcus species. The disease frequently affects the liver and lungs, which can lead to the development of hydatid cysts. These cysts may not exhibit any symptoms for years, which increases the risk of complications and delays in diagnosis. In order to manage the illness and avoid major consequences, early and accurate detection is crucial.

Serological methods have been shown to be useful supplemental tools in the diagnosis of echinococcosis when paired with radiological investigations. Enzyme-Linked Immunosorbent Assay (ELISA) is one of the most widely used immunodiagnostic methods due to its low cost, ease of use, and relatively high sensitivity. Other serological methods are immunoblotting, indirect hemagglutination assay, latex agglutination, and recombinant antigen-based assays, which have also enabled improved diagnostic methods.

The pathophysiology, epidemiology, and serological diagnosis of human echinococcosis are summarized in this review, with a focus on ELISA-based detection techniques. The benefits, drawbacks, sensitivity, specificity, and most recent developments in immunodiagnostic methods are also covered in this review. It also emphasizes how crucial it is to integrate imaging modalities with serological results in order to make an accurate diagnosis.

All things considered, especially in endemic areas, serological techniques are still crucial for the early identification and clinical treatment of echinococcosis. Future disease surveillance, treatment results, and preventative measures may all be enhanced by ongoing research and development of more sensitive and accurate diagnostic instruments.

 Modern enterprises operate in highly dynamic environments where decision-making depends on the seamless integration of presentation systems, analytics platforms, and operational intelligence tools. Despite the availability of advanced software ecosystems such as Microsoft PowerPoint, Power BI, and Redzone, organizations often struggle to unify narrative communication, real-time analytics, and shop-floor operational visibility into a single coherent decision-support framework. This paper proposes an integrated enterprise intelligence framework that connects presentation intelligence, business analytics, and operational execution layers. The objective is to enhance decision speed, accuracy, and traceability across organizational hierarchies.

The framework leverages Microsoft PowerPoint as a structured narrative layer for executive communication, Power BI as a data-driven analytical engine, and Redzone as an operational intelligence system for frontline workforce performance monitoring. Drawing insights from data hiding techniques, optimization methods, and intelligent system design principles (Bender et al., 1996; Gielen and Rutenbar, 2000; Clerc, 2002), the proposed model ensures that insights are not only generated but effectively communicated and operationalized.

The paper also explores how advanced computational methods such as particle swarm optimization (Kennedy and Eberhart, 1995) and statistical optimization (Medeiro et al., 1994) can enhance analytics pipelines within Power BI, while structured communication models inspired by information hiding and watermarking techniques (Petitcolas et al., 1999; Johnson et al., 2001) improve data integrity in enterprise reporting. The study concludes that integrated intelligence systems significantly improve enterprise agility, reduce decision latency, and enhance operational efficiency.

This study presents a 25-year spatiotemporal assessment of vegetation dynamics and drought severity in the Khorezm irrigated zone of the Amu Darya Basin (59–62.5°E, 40.5–42.5°N) over the period 2000–2024. Annual Normalized Difference Vegetation Index (NDVI) values were derived from Landsat Collection 2 Surface Reflectance imagery accessed via the Microsoft Planetary Computer STAC API, filtered to cloud cover below 20% and growing season months (April–September). The Vegetation Condition Index (VCI) was computed to classify annual drought severity according to the five-class Kogan framework. The Mann-Kendall non-parametric trend test and Sen's slope estimator were applied to assess long-term vegetation trends. Results indicate a mean annual NDVI of 0.1094 across the study period, with a minimum of 0.0754 in 2001 (Extreme drought) and a maximum of 0.1442 in 2022 (No drought). VCI classification identifies two years of Extreme drought (2000–2001), one year of Moderate drought (2008), fourteen years of Mild drought (56%), and eight years of No drought (32%). The Mann-Kendall test detects a positive trend direction (Kendall's τ = +0.267, p = 0.065, Sen's slope = +0.000548 NDVI yr⁻¹) that does not reach statistical significance at p < 0.05, reflecting high interannual variability and non-monotonic dynamics. The findings characterize the Khorezm irrigated zone as a landscape under chronic structural water deficit, with the 2016–2022 period representing the best sustained vegetation conditions of the entire record.

Large-scale data analysis has become a critical requirement in modern business intelligence systems due to exponential growth in structured and semi-structured data. Traditional spreadsheet tools are often insufficient for handling massive datasets efficiently, leading to performance bottlenecks, delayed computations, and limited scalability. Microsoft Excel Power Pivot, combined with in-memory data modeling, provides a powerful solution for optimizing performance in analytical environments. It enables efficient data compression, faster query execution, and advanced analytical capabilities through the VertiPaq engine and relational data modeling structures. This technical paper explores the architecture, optimization strategies, and performance enhancement techniques associated with Power Pivot, focusing on its integration with business intelligence workflows. It also discusses real-world applications in enterprise reporting, predictive analytics, and decision support systems. The study highlights how in-memory processing significantly reduces computational overhead and enhances scalability compared to traditional disk-based systems. Furthermore, comparisons are drawn with distributed computing frameworks such as MapReduce and Spark to contextualize the role of Excel-based modeling in modern analytics ecosystems. The findings suggest that Power Pivot, when properly optimized, serves as a highly efficient tool for medium to large-scale analytical workloads, bridging the gap between end-user analytics and enterprise-level data processing systems.