The Federal University of Technology Akure (informally FUT Akure or simply FUTA) was founded in 1981 under a drive by the government of Nigeria to create universities that specialised in producing graduates with practical as well as theoretical knowledge of technologies. It is located in Akure, the capital city of Ondo State.Other universities of technology established around the same time were the Federal University of Technology Owerri, the Federal University of Technology, Abeokuta (FUTAB), which later metamorphosed to the present University of Agriculture, Abeokuta (FUNAAB), the Federal University of Technology Minna, the Federal University of Technology Yola, and the Federal University of Technology Bauchi (now Abubakar Tafawa Balewa University).
Epilepsy, a chronic neurological disorder characterized by recurrent seizures, affects approximately 1
This paper introduces and analyzes a novel hybrid reaction–diffusion model that integrates key nonlinear features from the BVAM, Schnakenberg, and Gray–Scott systems. The proposed model captures both local activation and long-range inhibition through a complex interplay of autocatalytic and cross-inhibitory kinetics. We perform a detailed mathematical analysis including linear stability, Turing instability criteria, and Lyapunov exponent computations to characterize the onset of spatiotemporal complexity. Numerical simulations in both one and two spatial dimensions reveal a diverse spectrum of dynamic behaviors, ranging from Turing-type stationary patterns (spots, stripes, and labyrinths) to high-dimensional chaotic oscillations. The presence of positive Lyapunov exponents and a non-integer Kaplan–Yorke dimension confirms the emergence of deterministic chaos in both time and space. All simulations were conducted using a custom MATLAB implementation based on the Split-Step Fourier Method (SSFM). The results establish the hybrid model as a unifying framework capable of capturing rich pattern formation dynamics in nonlinear systems, with potential applications in developmental biology, chemical media, and ecological systems. This work extends classical reaction-diffusion theory by demonstrating how hybridized kinetics can lead to robust and tunable spatiotemporal chaos.
Abrasive Index (AI) is crucial in evaluating coal’s impact on mechanical equipment wear, influencing operational efficiency and maintenance costs. Accurate AI prediction is essential for optimizing coal utilization and reducing economic losses in industrial applications. This study systematically evaluates AI prediction using 18 supervised learning (SL) models on a dataset of 129 coal samples from the Witbank Coalfield, South Africa. The most suitable model is identified based on performance comparisons, followed by hyperparameter optimization using eight metaheuristic optimization algorithms (MOAs). To further ensure robustness and interpretability, Monte Carlo Simulation (MCS), SHapley Additive exPlanations (SHAP), and Individual Conditional Expectation (ICE) methods are employed for uncertainty quantification and feature importance analysis. The results indicate that the Dung Beetle Optimizer-optimized Random Forest (DBO-RF) model outperforms all other hybrid models, achieving an R2 of 0.94, RMSE of 26.583, MAPE of 0.202, and VAF of 94.178% on the test set. Feature interpretation analysis reveals that quartz content significantly impacts AI predictions, while volatile matter contributes the least. This study provides a comprehensive AI prediction framework that integrates SL, optimization, and interpretability analysis, offering valuable insights for coal selection, equipment maintenance, and industrial process optimization.
The higher heating value (HHV) and proximate analysis of coal are widely recognized as critical indices for coal quality across various applications. Given the high cost and labor-intensive nature of experimental HHV determination, different models have been developed for rapid assessment. However, the performances of these models are often constrained by site-specific characteristics, limiting their applicability in new or untested coalfields. To address this limitation, this study aims to enhance the extrapolation potential of HHV prediction models through the development of a novel grey relational analysis-based committee of machine learning models (GRA-CMLM). Initially, three individual machine learning models were trained using laboratory-generated datasets and demonstrated excellent predictive performance. The artificial neural network (ANN) model achieved the highest R2 values of 0.9894, 0.9904, and 0.9667, for the training, testing, and validation phases respectively, followed by multiple linear regression (MLR) with R2 values of 0.9763, 0.9771, and 0.9348, and support vector regression (SVR) with 0.9752, 0.9766, and 0.9428. When applied to external datasets from an Indian coalfield, the models performed less effectively, with ANN, MLR, and SVR yielding R2 values of 0.5642, 0.7634, and 0.7339, respectively, highlighting the limitation in their transferability compared to results from South African coalfields. However, the GRA-based CMLM significantly improved prediction accuracy, achieving R2 value of 0.902 on the same external dataset. This demonstrates enhanced generalization and predictive performance over individual models. Furthermore, sensitivity analysis using the GRA method identified Volatile Matter (VM) as the most influential parameter among the proximate analysis variables in model development.
Epilepsy is a chronic neurological disorder that significantly affects the quality of life (QoL), often causing irreversible brain damage and physical impairment. Electroencephalography (EEG) signal analysis is crucial for monitoring epilepsy, enabling early seizure detection and timely intervention. Effective seizure detection requires identifying interpretable features from the EEG signal to improve clinical outcomes. This study proposes a novel interpretable multi-view feature learning approach (IMV-FL), in which the time-domain signals and Discrete Fourier Transform (DFT) are applied to convert the time-domain EEG signal into frequency-domain representations. To develop initial multiview feature extraction and compression, spatial and temporal morphological features are extracted from optimal layers of ResNet and Long Short-Term Memory (LSTM) models, with feature compression performed using a Deep Neural Network (DNN). To construct an interpretable multi-view feature fusion, linear and nonlinear properties are calculated for the features and for the fusion strategies. The selected features are processed using the Mutual Information-Based Feature (MIBF) selection algorithm, and a Stacking Ensemble Classifier (SAEC) is adopted for unified-view classification. To enhance clinical interpretability, SHapley Additive exPlanations (SHAP) is applied. The proposed framework outperforms single-view feature learning methods by 3% on average and state-of-the-art techniques by 2% in classification accuracy, sensitivity, specificity, and F1-score using the CHB-MIT Scalp and Bonn EEG datasets. This approach offers an effective tool for EEG-based seizure detection (ESD) in clinical and healthcare settings.
