Estimating flexible Weibull extension models under progressive Type-II censoring presents a statistical challenge due to the truncation of data at varying time points. This study proposes a parametric approach to estimate these models, accommodating the unique censoring mechanism. The flexible Weibull extension allows for a versatile representation of survival data, incorporating shape parameters that adapt to different hazard functions. Methodologically, maximum likelihood estimation is employed, leveraging the observed data and the censoring information effectively. The study explores the theoretical foundations and practical implications of this estimation technique, highlighting its applicability in analysing survival times subject to progressive Type-II censoring. Simulation studies are conducted to assess the performance of the proposed method under various scenarios, demonstrating its robustness and accuracy. Finally, real-world data examples illustrate the utility of the approach in modeling and interpreting survival data affected by progressive Type-II censoring, thereby contributing to advancements in statistical survival analysis methodologies.

Credit card fraud poses significant challenges to financial institutions and consumers worldwide, necessitating robust and effective detection mechanisms. This paper explores the application of machine learning and data science techniques to detect fraudulent activities in credit card transactions. By leveraging advanced algorithms, including supervised and unsupervised learning methods, we aim to identify patterns and anomalies indicative of fraud. We discuss the pre-processing steps, feature engineering, and model selection processes critical to developing an accurate and efficient detection system. Our results demonstrate the potential of machine learning models, such as decision trees, random forests, and neural networks, to enhance fraud detection capabilities, reduce false positives, and protect against financial losses. This research underscores the importance of integrating data science techniques in the fight against credit card fraud, offering insights into future directions and improvements in fraud detection methodologies.

This article analyzes the topology of telecommunication networks (TN) with various structures and reveals that, despite the numerous ways to connect nodes, network construction is based on three main (basic) topologies: bus, ring, and star. It is shown that in telecommunication networks, radial, ring, and radial-ring structures are among the most frequently used typical structures. The penalty function method determines the optimal TN structure based on the criterion of overall network reduced costs. Additionally, when choosing the most preferred basic typical structure for building telecommunication networks, the relative efficiency coefficient is proposed as an indicator.

Helm charts function as the principal subject of examination in this study regarding their role in simplifying Kubernetes application deployment and management procedures. Developers find it difficult to deploy Kubernetes applications through container orchestration even though Kubernetes offers scalable platforms and fault tolerance due to its complex nature for those without configuration experience. Through its package manager role, Helm provides reusable charts known as "charts," which solve deployment issues by streamlining the installation process. Developers can automate Kubernetes deployments through Helm charts because these tools create standardized packages that organize application configurations together with their dependencies. An automated Helm package system removes difficult-to-handle YAML configuration files, establishing uniform deployments and minimizing mistakes from human operators. Helm charts give developers three main advantages: user-friendly deployment through charts alongside version management features and rollback protocols and automatic environment deployment configuration capabilities. The paper analyzes Helm's position in Kubernetes management through discussions about its CI/CD pipeline integration and GitOps workflow adoption. The paper examines future developments in Kubernetes and Helm mainly through the lens of enhanced multi-cluster management and serverless architecture growth. The paper supports Helm charts because they enable organizations to achieve faster deployment times, enhanced practice management, and decreased Kubernetes environment complexity. Helm charts within Kubernetes deployment methods produce simplified processes and grant application managers complete control to operate across different infrastructure types.

The rapid evolution of artificial intelligence, particularly large language models (LLMs), has profoundly influenced software engineering practices. Software architecture design, microservice migration, code generation, quality assurance, and risk analysis are increasingly supported by intelligent systems capable of analyzing and generating complex software artifacts. Despite the growing body of literature addressing artificial intelligence in software engineering, a comprehensive synthesis that integrates architectural decision support, microservice recommendation, automated code analysis, and generative AI-driven development processes remains limited. This research presents an extensive multivocal literature-based investigation into the role of machine learning and large language models in modern software architecture and microservice engineering. Drawing upon a wide range of empirical studies, systematic literature reviews, and foundational research in software engineering, the study examines how AI technologies are transforming architectural design processes, system modularization, requirements engineering, and software quality management.

The research employs a qualitative analytical methodology informed by systematic literature review practices and grounded theory-inspired synthesis to interpret findings across existing academic works. The analysis investigates the theoretical foundations, architectural implications, and practical applications of generative AI and machine learning techniques in software engineering environments. Particular emphasis is placed on microservice recommendation systems, migration from monolithic architectures, architectural pattern identification, code generation systems, and automated software quality assessment. Furthermore, the research explores the implications of integrating knowledge graphs with LLMs for improved knowledge access in architectural decision-making.

Findings reveal that AI-driven development tools are reshaping the software engineering lifecycle by enabling automated reasoning over codebases, facilitating design pattern discovery, improving risk analysis, and supporting architectural decision-making. However, the study also identifies significant challenges related to reliability, governance, explainability, and integration with mission-critical systems. The results highlight both the transformative potential and the critical limitations of current AI-driven software engineering practices. The paper concludes by outlining future research directions aimed at improving the reliability, transparency, and architectural integration of AI technologies within complex software ecosystems.