Vol. 12, No. 03 [March 2026]

A recap will be made on Bayesian inference models with unknown mean (mu) and variance ( 𝜎2). These models use conjugate prior distributions, commonly a Normal-Gamma or Inverse Normal-Gamma distribution, to jointly estimate the parameters and the model when the variance is unknown. A modification is made to find the same parameters jointly. This can be seen in the article "Fundamentals for Obtaining New Probability Distribution Functions with Bayesian Inference with Gaussian and Near Gaussian Behavior," which demonstrates the use of Bayesian inference with only the known variance and modifies the algorithm's input data to estimate the mean. As will be seen below, both methods yield the same result.
Keywords: Bayesian Inference, Normal pdf, t Student pdf, Likelihood, Inverse Normal-Gamma pdf

In this new article, the Adomian Decomposition Method is presented, along with the application of the ZJ transform together for the solution of nonlinear Mixed PDEs of order n with p+q =n, where n is a positive integer, with good effectiveness and potential of this hybrid approach to obtain approximate analytical solutions of nonlinear problems.
Keywords: ZJ Transform, Adomian Decomposition Method, Adomian Polynomials, Nonlinear Differential Equations, Series, Mixed Nonlinear Partial Differential Equations.

The concept of emotionally neutral design is examined as a key factor in the formation and maintenance of user trust in high-risk environments such as financial technologies and artificial intelligence-based systems. The relevance of the topic is shaped by the trust crisis of 2024–2025, driven by the growing complexity of algorithmic decision-making and the spread of manipulative interfaces. The aim of the study is to provide a theoretical rationale for, and a practical assessment of, the effectiveness of minimalist and restrained interfaces in reducing cognitive load and strengthening user agency. The methodological foundation combines a systematic literature review, a comparative analysis of Gartner and McKinsey market data, and an in-depth examination of case studies involving the deployment of personal financial management (PFM) systems with a user base exceeding 1.3 million active users. The empirical dataset analyzed in this study covers approximately twelve months of interaction logs collected between 2023 and 2024 from a large-scale fintech platform. This dataset enabled the observation of both short-term behavioral reactions and longer-term patterns of financial decision-making under different interface configurations.
The results indicate that the combination of emotional neutrality with the principles of Calm Technology contributes to higher user loyalty and an increase in retained funds, while 61% of users associate trust primarily with ease of navigation and the absence of visual pressure. The study confirms the hypothesis that, in high-risk products, design performs a strategic mediating function by narrowing the “agency gap” between the algorithm and the human user. The findings underscore the practical importance of shifting from emotional engagement toward architectural transparency, a transition of clear relevance for product designers, AI system architects, and technology executives seeking to build resilient and ethical digital ecosystems.
Keywords: product design, high-risk products, digital trust, emotionally neutral design, artificial intelligence, financial technologies, cognitive load, human-centered AI, user agency, calm technology.

The selection of electrical conductors plays a critical role in the design and implementation of electrical systems, directly affecting system reliability, efficiency, and safety. However, this process is inherently complex due to the diversity of available products with varying technical specifications and costs. This study proposes an integrated TOPSIS–Entropy approach to identify the optimal copper conductor among 28 alternatives provided by a supplier. Nine criteria are considered, including nominal cross-sectional area, number of strands, strand diameter, insulation thickness between strands, sheath thickness, overall diameter, maximum DC resistance at 20°C, mass per unit length, and cost. The Entropy method is employed to objectively determine the weights of the criteria, followed by the application of TOPSIS to rank the alternatives. The results demonstrate that product code 20255114 is the most suitable option.
Keywords: copper conductor selection, MCDM, TOPSIS, Entropy

Large-scale ETL (Extract, Transform, Load) pipelines sit at the heart of modern data infrastructure, yet ensuring their integrity and accuracy at scale remains a persistent challenge. This study develops and evaluates a multi-layered framework that integrates automated validation tools, model-driven development, and machine-learning-driven predictive analytics to address common failure points across the data lifecycle. Employing a qualitative and agile research design over a six-month period in the United States, the study iteratively tested ETL components under varying data loads. Key findings show that model-driven development significantly reduced transformation errors, automated validation tools enabled real-time error detection with measurable governance impact, and ML-based prediction prevented quality degradation before downstream effects materialized. Collectively, these innovations offer scalable, adaptive solutions for data-driven organizations seeking to strengthen decision-making through more reliable data pipelines.
Keywords: ETL, Data Engineering, Data Quality, Artificial Intelligence, Machine Learning, Data Governance, Frameworks, Data Transformation