.
This study proposes an enhanced method for detecting malicious nodes in wireless sensor networks (WSNs) by extending security measures based on the least-squares-support vector machine (LS-SVM). Specifically, a Pearson correlation coefficient-based enhanced LS-SVM (ELS-SVM) is employed for identifying malicious nodes. Detection of such nodes through machine learning (ML) algorithms establishes a robust defense mechanism, enabling the monitoring of node behavior and the mitigation of potential threats. In this context, effective anomaly detection in WSNs utilizing the ELS-SVM fosters a secure environment, further strengthened by transparent logging and secure data storage (SDS) via blockchain technology. SDS in WSNs is critical, and this research achieves it through a hybrid approach combining the ELS-SVM and blockchain technology for malicious node detection (MND). By ensuring network reliability and data immutability, this hybrid mechanism offers enhanced protection against potential attacks.
This paper presents a comprehensive investigation into the prediction of axial load capacity (P) for elliptical double steel columns (EDSCs) using a diverse set of machine learning models (MLMs). These include Artificial Neural Network (ANN), Gene Expression Programming (GEP), Support Vector Regression (SVR), Random Forest (RF), and AdaBoost. Among the models, AdaBoost demonstrated superior performance, achieving an R2 of 0.996 and a MAPE of 0.013 during training, outperforming other models under identical conditions. Using a dataset of 119 finite element models derived from prior experimental research, the study validates the proposed solution through k-fold cross-validation, feature importance analysis, and detailed comparisons with experimental data. A Graphical User Interface (GUI) was developed specifically for the AdaBoost model due to its superior accuracy and efficiency, offering engineers a practical and accessible tool for axial load prediction in EDSC design. This research highlights the significance of using advanced machine learning techniques for structural engineering applications, providing valuable insights for the optimization of EDSC performance and design under varying conditions.
This manuscript addresses the predefined synchronization in quaternion-valued neural networks (QVNNs) incorporating mixed time delays and leakage terms. Sufficient conditions for predefined-time synchronization of time-delayed QVNNs have been derived. The QVNNs have been decomposed into real-valued systems, and further synchronization criteria have been deliberated by constructing a suitable Lyapunov function. The novel aspect of the suggested scheme is that two new controllers are utilized to achieve predefined-time synchronization for quaternion-valued neural networks (QVNNs) with mixed and leakage time delays. Unlike traditional methods, this approach ensures synchronization within a specific time frame. The first controller uses the sign function, while the second employs exponential and Lyapunov functions, offering simplicity and fewer parameters. This method's flexibility and superior timing adjustment capabilities extend beyond traditional asymptotic finite-time and fixed-time synchronization, adding significant value to the literature by enhancing the practical utility of QVNNs in complex systems.
This study investigates the growing public interest in international scholarship programs among Pakistani students using Google Trends data from 2020 to 2025. Quantitative analysis: The online search behavior analysis used a quantitative approach to compare variations in eliciting online search behavior across five major scholarship programs: Fulbright, DAAD and Erasmus Mundus, Türkiye Bursları, and CSC. A quantitative descriptor was referred to define the relative search volume, mainly because the Elon Publicity indicator commences public awareness and aspirations regarding global higher education. As a result, overall search interest has consistently grown from 2018 to 2022, with seasonal peaks that closely match relevant program application deadlines. At the relative level, Fulbright and DAAD had high search interest among Turkish searches, which remained primarily high after 2021 for Erasmus Mundus students, driven by the 2021 rapid expansion in the number of destinations, a sign of diverse destinations. Regional approximation and an overall strong concentration of interest in Punjab, Islamabad, and Sindh are booming and topic-rich, with Khyber Pakhtunkhwa and Balochistan increasingly joining the topic. The results point to a shift from dissipative adaptation’ outwards academic mobility from PA “dissipative adaptation,’ to a valorization of “creative adaptations,” a kind of global academic engagement for the future, emphasizing the bridging role of international scholarships in knowledge exchange and co-creation.
Energy harvesting from waste heat can improve energy efficiency in society. This research investigated the structural behaviors of lead zirconate titanate–based ferroelectric ceramics using operando neutron diffraction measurements under the conditions of two energy-harvesting cycles that involve consideration of the temperature changes of automobile exhaust gas for achieving good harvesting efficiencies. Input and output electrical energies and neutron diffraction data were simultaneously collected. The obtained time-resolved neutron diffraction intensity data indicate that the applied electric fields and temperature changes induced 90° domain rotation and lattice strain. These structural changes and their variations depending on cycle conditions, such as temperature changes, applied electric fields, and circuit switching, provide insight into the origins of the differences in the behaviors of electrical input/output energies in the cycles.
