Machine ranking plays a crucial role in manufacturing, influencing both economic efficiency and technical performance. However, this task is inherently complex because each machine must be evaluated simultaneously across multiple, often conflicting, criteria. This challenge highlights the need for Multi-Criteria Decision-Making (MCDM) methods. This study proposes a new criteria weighting approach, the SITDER method, that aims to improve the stability of machine ranking results across different MCDM techniques. SITDER integrates two methods: Skewness Impact Through Distributional Evaluation (SITDE) and Rank Order Centroid (ROC). In the first stage, SITDE is used to determine the relative importance of the criteria. In the second stage, ROC is applied to compute the final criterion weights. The effectiveness of SITDER was validated using two case studies involving the ranking of 13 handheld polishers and 12 grinding machines. In both examples, the rank correlation coefficients among the MCDM methods were consistently higher when using SITDER compared to SITDE alone. In Example 1, the average Spearman coefficient increased from 0.6282 (SITDE) to 0.739 (SITDER). In Example 2, it increased from 0.9643 to 0.9744. These results demonstrate that SITDER is a highly effective weighting method and a valuable addition to the set of available criteria weighting techniques.

Identifying potential marsh areas in Iraq is essential for sustainable water resource monitoring in arid regions. This study uses altimetry, interferometric synthetic aperture radar techniques, and hydrological modeling for identifying potential marsh sites in the southwestern Iraq desert. In the study's methodology, altimetry data were used to provide precise Euphrates water elevations, while Interferometric Synthetic Aperture Radar (InSAR)-derived elevation enhanced the spatial detail of the Digital Elevation Model (DEM). A stream burning technique was applied to improve DEM-2024 and ensure accurate hydrological representation. This integration enabled the generation of a high-resolution topographic model capable of identifying low-lying zones with high potential for marsh restoration. Three DEMs from 2022, 2023, and 2024 were evaluated for accuracy using Real-Time Kinematic (RTK) points. DEM 2024, enhanced by the stream burning technique, was the most accurate, with a Root Mean Square Error (RMSE) of 3.27 m, a bias of -0.86 m, and a Standard Deviation (SD) of 3.15 m. This improved DEM facilitated the reliable identification of potential marsh boundaries. The Euphrates River discharge value at 31.3351°N, 45.0527°E was 549.95 m3/s, which was the basis for hydrological simulations. Moreover, the channel characteristics (64.7 m width, 5.0 m depth, and 1.7 m/s velocity) validated the viability of water diversion for marsh creation. Geomorphological assessments identified basin 280 as the most suitable site for creating a marsh. With a total area of 725.74 km2, river length of 301,647.5 m, elongation ratio of 0.589, and circularity ratio of 0.650, it indicates favorable hydrological retention and flow distribution characteristics. This study demonstrates the integration of remotely sensed altimetry techniques, DEM optimization, and hydrological analysis to support large-scale marsh identification and restoration.

Model-driven engineering is significant in modern software development, as it leverages models to guide software design and implementation. However, legacy systems are characterized by outdated architectures and poor documentation, which pose challenges for maintenance. Model-Driven Reverse Engineering (MDRE) addresses these issues by extracting models from legacy systems, enabling better comprehension. This study explores clustering techniques to extract high-level concepts from source code. Clustering reveals system concepts that can be transformed into models by grouping similar code entities. These models provide a foundation for system modernization and further development. Unlike existing MDRE approaches, this study systematically evaluates multiple clustering techniques and preprocessing scenarios to assess their ability to recover high-level system concepts from legacy source code. The results highlight the impact of preprocessing and noise on concept extraction, providing insights for applying clustering within model-driven modernization workflows.

The impact of climate change may pose a threat to the magnitude and occurrence of river flow. Efficient watershed management requires a rational approach to predicting future climate change impacts. The Nith River in Ontario, Canada has experienced frequent flooding as a result of extreme storm events every year or sometimes multiple times per year. The purpose of this study is to provide guidance on building watershed resiliency through flooding condition assessments and the use of climate scenarios for the watershed. Three future climate scenarios are assessed, to identify the most suitable one. The North American Regional Climate Change Assessment Program (NARCCAP) provided a scenario using Canadian Regional Climate Model and Canadian Global Climate Model version 3 (CRCM-CGCM3); another scenario involved Regional Climate Model version 3 and CGCM3 (RCM3-CGCM3), while the Ministry of Natural Resources (MNR) provided an additional scenario. This statistical analysis represents the best match between the observed and projected data of MNR. According to future climatic data, precipitation-runoff was calculated by a hydrological model which is known as the Soil and Water Assessment Tool (SWAT). Calibration and validation were performed to optimize the model's results. Bias factors were calculated from the observed period data and used to adjust future runoff estimates. This research showed significant changes in storm magnitude and frequency due to climate change, providing insights for watershed management and development. However, the results may vary, if different climate scenarios and precipitation-runoff models are used.

Sperm morphology analysis is a key parameter in male fertility evaluation, but the manual process is subjective and time-consuming. Therefore, an automated deep learning classification system offers a potential solution. This study evaluates and compares MobileNetV2 and EfficientNetB0, individually and in combination via feature fusion, for sperm morphology image classification using the SMID dataset of 3000 images. MobileNetV2 and EfficientNetB0 models achieved accuracies of 62.75% and 33.06%, respectively, after 50 epochs, while the feature fusion model reached 85.04% in only 5 epochs. Thus, combining features from both architectures yields superior accuracy and efficiency for automated sperm analysis.

This research addresses the limitations of conventional, static survival models for Remaining Useful Life (RUL) estimation in Prognostics and Health Management (PHM) by proposing and implementing a data-driven framework. The study posits that a system's characteristic life, represented by the Weibull scale parameter (λ), is not static but a dynamic variable influenced by time-varying operational and environmental factors, or covariates. Unlike traditional methods, such as the Accelerated Failure Time (AFT) model, which rely on restrictive linear assumptions to link covariates to the lifetime model, this methodology employs supervised Machine Learning (ML) models to quantify this complex and potentially non-linear relationship. The methodology is validated using the National Aeronautics and Space Administration (NASA) Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) aircraft engine degradation dataset. The results demonstrate that ensemble-based methods like Light Gradient Boosting Machine (LightGBM) and Extreme Gradient Boosting (XGBoost) significantly outperform classical statistical models across a suite of standard regression metrics and custom-developed Weibull-centric error metrics. The consistent performance rankings across all metrics validate the effectiveness of the proposed approach and confirm its capability to capture the intricate degradation dynamics of real-world systems.

Accurate prediction of energy consumption in smart homes is essential for optimizing energy usage, reducing operational costs, and enhancing occupant comfort. Traditional statistical and machine learning models often struggle to capture the complex spatiotemporal patterns inherent in energy usage data. To address this limitation, this study proposes a hybrid Convolutional Neural Network–Long Short-Term Memory (CNN-LSTM) model for forecasting energy consumption in smart homes, combining the spatial feature extraction strengths of Convolutional Neural Networks (CNNs) with the temporal learning capabilities of Long Short-Term Memory (LSTM) networks. The model demonstrated high predictive accuracy and generalization, as reflected by the evaluation metrics: Mean Absolute Error (MAE) = 0.0325, Mean Squared Error (MSE) = 0.0031, Root Mean Squared Error (RMSE) = 0.0557, and Mean Absolute Percentage Error (MAPE) = 0.2054. However, as the model was trained on synthetic data, further validation using real-world energy consumption datasets is necessary to assess its practical applicability and robustness.

The Knowledge of long-term energy demand, is essential for sustainable energy planning, especially in emerging economies. This study examines the relationship between Final Energy Consumption (FEC) and eleven economic variables in Malaysia from 1980 to 2022. The analysis used univariate regressions, feature selection techniques (forward selection, lasso, and recursive feature elimination), and multiple regression with diagnostic checks to systematically evaluate the explanatory power of each predictor. The results reveal that Gross Domestic Product (GDP), manufacturing output, electricity consumption, and urban population are the first factors influencing FEC, while imports, exports, and Foreign Direct Investment (FDI) play secondary roles. Population and trade balance exhibit redundancy and limited predictive value. Diagnostic testing, including the Variance Inflation Factor (VIF), Durbin–Watson statistics, and residual analysis, confirms the reliability of the regression framework. The results show that GDP, manufacturing output, and urban population are the most significant predictors of long-term energy demand (R² = 0.94, p < 0.01), whereas trade-related variables have a moderate impact. These findings underscore the importance of aligning industrial growth, urbanization policies, and trade strategies with sustainable energy planning, highlighting the advantages of using regression with feature selection and diagnostics.

Concrete exposed to high temperatures undergoes severe deterioration in strength and microstructural integrity, demanding effective post-fire rehabilitation. This study investigates Magnetized Water (MW) as a re-curing medium for restoring heat-damaged Self-Compacting Concrete (SCC). Specimens prepared with either MW or Normal Water (NW) were subjected to 200°C, 400°C, and 600°C, then re-cured for 28 days in magnetized or plain water. Performance recovery was assessed through compressive, flexural, and bond strength tests, complemented by porosity and Ultrasonic Pulse Velocity (UPV) measurements. The results showed that MW significantly improved recovery compared with NW: after 600°C exposure, re-cured SCC regained 75%–80% of its original compressive strength versus 60%–65% with NW, with a residual strength consistently higher by 5%–15% across all temperatures. The flexural and bond strength recovery exhibited similar trends. MW also reduced porosity by 5%–10% and increased UPV, indicating a denser matrix. Overall, MW re-curing proved an efficient technique for restoring the mechanical properties of thermally deteriorated concrete, highlighting its potential as a practical strategy for post-fire rehabilitation and durability enhancement.

The rapid progression of Internet connections has led to a significant increase in cyber-attack events, often resulting in severe and disastrous consequences. Malware is one of the primary weapons used to achieve malicious objectives in cyberspace, either through the exploitation of newly discovered vulnerabilities or through the misuse of features introduced by new technologies. The development of more effective and robust malware defense mechanisms is considered a critical necessity in cybersecurity. Human–Computer Interface (HCI) systems have become increasingly important, and as communication, computing, and display technologies continue to advance, conventional HCI methods may become a bottleneck in effectively handling the growing data flow. Federated Learning (FL) is a Machine Learning (ML) paradigm that aims to train models through decentralized devices, whereas the local data are kept private. In this manuscript, a Secure and Efficient Federated Learning using Optimization Algorithms and Deep Learning for Cybersecurity Applications (SEFL-OADLCA) model is proposed for HCI systems. The aim is to present an effective FL framework to address cybersecurity issues. First, the min–max scaler method is applied for data preprocessing. For the Dimensionality Reduction (DR) process, the Mountain Gazelle Optimizer (MGO) technique is employed. Furthermore, a hybrid Temporal Convolutional Network and Long Short-Term Memory (TCN+LSTM) technique is utilized for the attack classification process. Finally, the hyperparameter selection process is performed using the Remora Optimization Algorithm (ROA) to optimize the classification results. The comparison study of the SEFL-OADLCA method portrayed a superior accuracy of 99.46% compared with existing approaches on the NSL-KDD dataset.