Search Results (2,620)

Zero-day attacks pose a persistent threat to computing infrastructure by exploiting previously unknown software vulnerabilities that evade traditional signature-based network intrusion detection systems (NIDSs). To address this limitation, machine learning (ML) techniques offer a promising approach for enhancing anomaly detection in network traffic. This study evaluates several ML models on a labeled network traffic dataset, with a focus on zero-day attack detection. Ensemble learning methods, particularly eXtreme gradient boosting (XGBoost), achieved perfect classification, identifying all 6231 zero-day instances without false positives and maintaining efficient training and prediction times. While classical support vector machines (SVMs) performed modestly at 64% accuracy, their performance improved to 98% with the use of the borderline synthetic minority oversampling technique (SMOTE) and SMOTE + edited nearest neighbours (SMOTEENN). To explore quantum-enhanced alternatives, a quantum SVM (QSVM) is implemented using three-qubit and four-qubit quantum circuits simulated on the aer_simulator_statevector. The QSVM achieved high accuracy (99.89%) and strong F1-scores (98.95%), indicating that nonlinear quantum feature maps (QFMs) can increase sensitivity to zero-day exploit patterns. Unlike prior work that applies standard quantum kernels, this study introduces a parameterised quantum feature encoding scheme, where each classical feature is mapped using a nonlinear function tuned by a set of learnable parameters. Additionally, a sparse entanglement topology is derived from mutual information between features, ensuring a compact and data-adaptive quantum circuit that aligns with the resource constraints of noisy intermediate-scale quantum (NISQ) devices. Our contribution lies in formalising a quantum circuit design that enables scalable, expressive, and generalisable quantum architectures tailored for zero-day attack detection. This extends beyond conventional usage of QSVMs by offering a principled approach to quantum circuit construction for cybersecurity. While these findings are obtained via noiseless simulation, they provide a theoretical proof of concept for the viability of quantum ML (QML) in network security. Future work should target real quantum hardware execution and adaptive sampling techniques to assess robustness under decoherence, gate errors, and dynamic threat environments. Full article

To address the issues of reduced operational efficiency, shortened equipment lifespan, and significant safety hazards caused by bearing wear and blade cavitation in hydroelectric turbine generators due to prolonged high-load operation, this paper proposes a ResNet + self-attention-based acoustic fingerprint fault diagnosis algorithm for hydroelectric turbine generators. First, to address the issue of severe noise interference in acoustic signature signals, the ensemble empirical mode decomposition (EEMD) is employed to decompose the original signal into multiple intrinsic mode function (IMF) components. By calculating the correlation coefficients between each IMF component and the original signal, effective components are selected while noise components are removed to enhance the signal-to-noise ratio; Second, a fault identification network based on ResNet + self-attention fusion is constructed. The residual structure of ResNet is used to extract features from the acoustic signature signal, while the self-attention mechanism is introduced to focus the model on fault-sensitive regions, thereby enhancing feature representation capabilities. Finally, to address the challenge of model hyperparameter optimization, a Bayesian optimization algorithm is employed to accelerate model convergence and improve diagnostic performance. Experiments were conducted in the real working environment of a pumped-storage power station in Zhejiang Province, China. The results show that the algorithm significantly outperforms traditional methods in both single-fault and mixed-fault identification, achieving a fault identification accuracy rate of 99.4% on the test set. It maintains high accuracy even in real-world scenarios with superimposed noise and environmental sounds, fully validating its generalization capability and interference resistance, and providing effective technical support for the intelligent maintenance of hydroelectric generator units. Full article

Palmprint recognition is becoming more and more common in the fields of security authentication, mobile payment, and crime detection. Aiming at the problem of small sample size and low recognition rate of palmprint, a small-sample palmprint recognition method based on image expansion and Dynamic Model-Agnostic Meta-Learning (DMAML) is proposed. In terms of data augmentation, a multi-connected conditional generative network is designed for generating palmprints; the network is trained using a gradient-penalized hybrid loss function and a dual time-scale update rule to help the model converge stably, and the trained network is used to generate an expanded dataset of palmprints. On this basis, the palmprint feature extraction network is designed considering the frequency domain and residual inspiration to extract the palmprint feature information. The DMAML training method of the network is investigated, which establishes a multistep loss list for query ensemble loss in the inner loop. It dynamically adjusts the learning rate of the outer loop by using a combination of gradient preheating and a cosine annealing strategy in the outer loop. The experimental results show that the palmprint dataset expansion method in this paper can effectively improve the training efficiency of the palmprint recognition model, evaluated on the Tongji dataset in an N-way K-shot setting, our proposed method achieves an accuracy of 94.62% ± 0.06% in the 5-way 1-shot task and 87.52% ± 0.29% in the 10-way 1-shot task, significantly outperforming ProtoNets (90.57% ± 0.65% and 81.15% ± 0.50%, respectively). Under the 5-way 1-shot condition, there was a 4.05% improvement, and under the 10-way 1-shot condition, there was a 6.37% improvement, demonstrating the effectiveness of our method. Full article

The discovery and design of high-entropy alloys (HEAs) faces significant challenges due to the vast combinatorial design space and uncertainties in thermodynamic data. This work presents a modular, uncertainty-aware computational framework with the primary objective of accelerating the discovery of solid-solution HEA candidates. The proposed pipeline integrates ensemble thermodynamic modeling, Monte Carlo-based estimation, and a structured three-phase pruning algorithm for efficient search space reduction. Key quantitative results are achieved in two main areas. First, for binary alloy thermodynamics, a Bayesian Neural Network (BNN) ensemble trained on domain-informed features predicts mixing enthalpies with high accuracy, yielding a mean absolute error (MAE) of 0.48 kJ/mol—substantially outperforming the classical Miedema model (MAE = 4.27 kJ/mol). These probabilistic predictions are propagated through Monte Carlo sampling to estimate multi-component thermodynamic descriptors, including $\Delta H_{\mathrm{mix}}$ and the $\mathrm{\Omega }$ parameter, while capturing predictive uncertainty. Second, in a case study on the Al-Cu-Fe-Ni-Ti system, the framework reduces a 2.4 million (2.4 M) candidate pool to just 91 high-confidence compositions. Final selection is guided by an uncertainty-aware viability metric, $P\left(\mathrm{HEA}\right)$, and supported by interpretable radar plot visualizations for multi-objective assessment. The results demonstrate the framework’s ability to combine physical priors, probabilistic modeling, and design heuristics into a data-efficient and interpretable pipeline for materials discovery. This establishes a foundation for future HEA optimization, dataset refinement, and adaptive experimental design under uncertainty. Full article

Emotion recognition in speech is essential for enhancing human–computer interaction (HCI) systems. Despite progress in Bangla speech emotion recognition, challenges remain, including low accuracy, speaker dependency, and poor generalization across emotional expressions. Previous approaches often rely on traditional machine learning or basic deep learning models, struggling with robustness and accuracy in noisy or varied data. In this study, we propose a novel multi-stream deep learning feature fusion approach for Bangla speech emotion recognition, addressing the limitations of existing methods. Our approach begins with various data augmentation techniques applied to the training dataset, enhancing the model’s robustness and generalization. We then extract a comprehensive set of handcrafted features, including Zero-Crossing Rate (ZCR), chromagram, spectral centroid, spectral roll-off, spectral contrast, spectral flatness, Mel-Frequency Cepstral Coefficients (MFCCs), Root Mean Square (RMS) energy, and Mel-spectrogram. Although these features are used as 1D numerical vectors, some of them are computed from time–frequency representations (e.g., chromagram, Mel-spectrogram) that can themselves be depicted as images, which is conceptually close to imaging-based analysis. These features capture key characteristics of the speech signal, providing valuable insights into the emotional content. Sequentially, we utilize a multi-stream deep learning architecture to automatically learn complex, hierarchical representations of the speech signal. This architecture consists of three distinct streams: the first stream uses 1D convolutional neural networks (1D CNNs), the second integrates 1D CNN with Long Short-Term Memory (LSTM), and the third combines 1D CNNs with bidirectional LSTM (Bi-LSTM). These models capture intricate emotional nuances that handcrafted features alone may not fully represent. For each of these models, we generate predicted scores and then employ ensemble learning with a soft voting technique to produce the final prediction. This fusion of handcrafted features, deep learning-derived features, and ensemble voting enhances the accuracy and robustness of emotion identification across multiple datasets. Our method demonstrates the effectiveness of combining various learning models to improve emotion recognition in Bangla speech, providing a more comprehensive solution compared with existing methods. We utilize three primary datasets—SUBESCO, BanglaSER, and a merged version of both—as well as two external datasets, RAVDESS and EMODB, to assess the performance of our models. Our method achieves impressive results with accuracies of 92.90%, 85.20%, 90.63%, 67.71%, and 69.25% for the SUBESCO, BanglaSER, merged SUBESCO and BanglaSER, RAVDESS, and EMODB datasets, respectively. These results demonstrate the effectiveness of combining handcrafted features with deep learning-based features through ensemble learning for robust emotion recognition in Bangla speech. Full article

Eye-written digit recognition presents a promising alternative communication method for individuals affected by amyotrophic lateral sclerosis. However, the development of robust models in this field is limited by the availability of datasets, due to the complex and unstable procedure of collecting eye-written samples. Previous work has proposed both conventional techniques and deep neural networks to classify eye-written digits, achieving moderate to high accuracy with variability across runs. In this study, we explore the potential of quantum machine learning by presenting a hybrid quantum–classical model that integrates a variational quantum circuit into a classical deep neural network architecture. While classical models already achieve strong performance, this work examines the potential of quantum-enhanced models to achieve such performance with fewer parameters and greater expressive capacity. To further improve robustness and stability, we employ an ensemble strategy that aggregates predictions from multiple trained instances of the hybrid model. This study serves as a proof-of-concept to evaluate the feasibility of incorporating a compact 4-qubit quantum circuit within a lightweight hybrid model. The proposed model achieves 98.52% accuracy with a standard deviation of 1.99, supporting the potential of combining quantum and classical computing for assistive communication technologies and encouraging further research in quantum biosignal interpretation and human–computer interaction. Full article

Diabetic retinopathy (DR), a leading cause of vision loss worldwide, poses a critical challenge to healthcare systems due to its silent progression and the reliance on labor-intensive, subjective manual screening by ophthalmologists, especially amid a global shortage of eye care specialists. Addressing the pressing need for scalable, objective, and interpretable diagnostic tools, this work introduces RetinoDeep—deep learning frameworks integrating hybrid architectures and explainable AI to enhance the automated detection and classification of DR across seven severity levels. Specifically, we propose four novel models: an EfficientNetB0 combined with an SPCL transformer for robust global feature extraction; a ResNet50 ensembled with Bi-LSTM to synergize spatial and sequential learning; a Bi-LSTM optimized through genetic algorithms for hyperparameter tuning; and a Bi-LSTM with SHAP explainability to enhance model transparency and clinical trustworthiness. The models were trained and evaluated on a curated dataset of 757 retinal fundus images, augmented to improve generalization, and benchmarked against state-of-the-art baselines (including EfficientNetB0, Hybrid Bi-LSTM with EfficientNetB0, Hybrid Bi-GRU with EfficientNetB0, ResNet with filter enhancements, Bi-LSTM optimized using Random Search Algorithm (RSA), Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), and a standard Convolutional Neural Network (CNN)), using metrics such as accuracy, F1-score, and precision. Notably, the Bi-LSTM with Particle Swarm Optimization (PSO) outperformed other configurations, achieving superior stability and generalization, while SHAP visualizations confirmed alignment between learned features and key retinal biomarkers, reinforcing the system’s interpretability. By combining cutting-edge neural architectures, advanced optimization, and explainable AI, this work sets a new standard for DR screening systems, promising not only improved diagnostic performance but also potential integration into real-world clinical workflows. Full article

Variations in global surface temperatures serve as critical indicators of climate change, and making accurate predictions regarding these patterns is essential for designing effective mitigation strategies. This study utilized a time series prediction methodology, leveraging annual temperature anomaly records from 1880 to 2022 provided by NASA’s GISTEMP v4 dataset. Following an extensive preprocessing phase, multiple deep learning models, namely, LSTM, DNN, CNN, and Transformer, were trained and analyzed separately. The individual model outputs were subsequently combined using a weighted averaging strategy grounded in linear regression, forming a novel LSTM and Transformer-based hybrid forecasting model. Model performance was assessed through widely recognized metrics including MSE, MAE, RMSE, and R². By integrating the distinct advantages of each model, the ensemble framework aimed to improve the overall predictive capability. The findings revealed that this hybrid design delivered more stable and accurate forecasts compared to any single model. The integration of diverse neural network structures proved effective in boosting predictive reliability and underscored the viability of deep learning-based hybrid modeling for climate trend forecasting. Full article

Considering the problems of having insufficient fault identification from single information sources in actual industrial environments, and different information sensitivity in multi-information source data, and different sensitivity of artificial feature extraction, which can lead to difficulties of effective fusion of equipment information, insufficient state representation ability, low fault identification accuracy, and poor robustness, a multi-information fusion fault identification network model based on deep ensemble learning is proposed. The network is composed of multiple sub-feature extraction units and feature fusion units. Firstly, the fault feature mapping information of each information source is extracted and stored in different sub-models, and then, the features of each sub-model are fused by the feature fusion unit. Finally, the fault recognition results are obtained. The effectiveness of the proposed method is evaluated by using two gearbox datasets. Compared with the method of simple stacking fusion and single measuring point without fusion, the accuracy of each type of fault recognition of the proposed method is close to 100%. The results show that the proposed method is feasible and effective in the application of gearbox fault recognition. Full article

This article presents a novel method for detecting series arc faults (SAFs) in residential applications using time-domain imaging (TDI) and Long Short-Term Memory (LSTM) networks. The proposed method transforms current signals into grayscale images by filtering out the fundamental frequency, allowing key arc fault characteristics—such as high-frequency noise and waveform distortions—to become visually apparent. The use of Ensemble Empirical Mode Decomposition (EEMD) helped isolate meaningful signal components, although it was computationally intensive. To address real-time requirements, a simpler yet effective TDI method was developed for generating 2D images from current data. These images were then used as inputs to an LSTM network, which captures temporal dependencies and classifies both arc faults and appliance types. The proposed TDI-LSTM model was trained and tested on 7000 labeled datasets across five common household appliances. The experimental results show an average detection accuracy of 98.1%, with reduced accuracy for loads using thyristors (e.g., dimmers). The method is robust across different appliance types and conditions; comparisons with prior methods indicate that the proposed TDI-LSTM approach offers superior accuracy and broader applicability. Trade-offs in sampling rates and hardware implementation were discussed to balance accuracy and system cost. Overall, the TDI-LSTM approach offers a highly accurate, efficient, and scalable solution for series arc fault detection in smart home systems. Full article

Urban multimodal travel trajectory prediction is a core challenge in Intelligent Transportation Systems (ITSs). It requires modeling both spatiotemporal dependencies and dynamic interactions among different travel modes such as taxi, bike-sharing, and buses. To address the limitations of existing methods in capturing these diverse trajectory characteristics, we propose a spatiotemporal multi-model ensemble framework, which is an ensemble model called GLEN (GCN and LSTM Ensemble Network). Firstly, the trajectory feature adaptive driven model selection mechanism classifies trajectories into dynamic travel and fixed-route scenarios. Secondly, we use a Graph Convolutional Network (GCN) to capture dynamic travel patterns and Long Short-Term Memory (LSTM) network to model fixed-route patterns. Subsequently the outputs of these models are dynamically weighted, integrated, and fused over a spatiotemporal grid to produce accurate forecasts of urban total traffic flow at multiple future time steps. Finally, experimental validation using Beijing’s Chaoyang district datasets demonstrates that our framework effectively captures spatiotemporal and interactive characteristics between multimodal travel trajectories and outperforms mainstream baselines, thereby offering robust support for urban traffic management and planning. Full article

With the continued growth in wind power installed capacity and electricity generation, accurate wind power forecasting has become increasingly critical for power system stability and economic operations. Currently, short-term wind power forecasting often employs deep learning models following modal decomposition of wind power time series. However, the optimal length of the time series used for decomposition remains unclear. To address this issue, this paper proposes a short-term wind power forecasting method that integrates improved modal decomposition with deep learning techniques. First, the historical wind power series is segmented using the Pruned Exact Linear Time (PELT) method. Next, the segmented series is decomposed using an enhanced Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (ICEEMDAN) to extract multiple modal components. High-frequency oscillatory components are then further decomposed using Variational Mode Decomposition (VMD), and the resulting modes are clustered using the K-means algorithm. The reconstructed components are subsequently input into a Long Short-Term Memory (LSTM) network for prediction, and the final forecast is obtained by aggregating the outputs of the individual modes. The proposed method is validated using historical wind power data from a wind farm. Experimental results demonstrate that this approach enhances forecasting accuracy, supports grid power balance, and increases the economic benefits for wind farm operators in electricity markets. Full article

Background/Objectives: Clinicians routinely rely on periapical radiographs to identify root-end disease, but interpretation errors and inconsistent readings compromise diagnostic accuracy. We, therefore, developed an explainable, multimodal AI framework that (i) fuses two data modalities, deep CNN embeddings and radiomic texture descriptors that are extracted only from lesion-relevant pixels selected by Grad-CAM, and (ii) makes every prediction transparent through dual-layer explainability (pixel-level Grad-CAM heatmaps + feature-level SHAP values). Methods: A dataset of 2285 periapical radiographs was processed using six CNN architectures (EfficientNet-B1/B4/V2M/V2S, ResNet-50, Xception). For each image, a Grad-CAM heatmap generated from the penultimate layer of the CNN was thresholded to create a binary mask that delineated the region most responsible for the network’s decision. Radiomic features (first-order, GLCM, GLRLM, GLDM, NGTDM, and shape2D) were then computed only within that mask, ensuring that handcrafted descriptors and learned embeddings referred to the same anatomic focus. The two feature streams were concatenated, optionally reduced by principal component analysis or SelectKBest, and fed to random forest or XGBoost classifiers; five-view test-time augmentation (TTA) was applied at inference. Pixel-level interpretability was provided by the original Grad-CAM, while SHAP quantified the contribution of each radiomic and deep feature to the final vote. Results: Raw CNNs achieved a ca. 52% accuracy and AUC values near 0.60. The multimodal fusion raised performance dramatically; the Xception + radiomics + random forest model achieved a 95.4% accuracy and an AUC of 0.9867, and adding TTA increased these to 96.3% and 0.9917, respectively. The top ensemble, Xception and EfficientNet-V2S fusion vectors classified with XGBoost under five-view TTA, reached a 97.16% accuracy and an AUC of 0.9914, with false-positive and false-negative rates of 4.6% and 0.9%, respectively. Grad-CAM heatmaps consistently highlighted periapical regions, while SHAP plots revealed that radiomic texture heterogeneity and high-level CNN features jointly contributed to correct classifications. Conclusions: By tightly integrating CNN embeddings, mask-targeted radiomics, and a two-tiered explainability stack (Grad-CAM + SHAP), the proposed system delivers state-of-the-art lesion detection and a transparent technique, addressing both accuracy and trust. Full article

Background: The red nucleus (RN) is a phylogenetically conserved structure within the midbrain that is traditionally associated with general motor coordination; however, its specific role in controlling directional movement remains poorly understood. Methods: This study systematically investigates the function and mechanism of RN neurons in directional movement by combining stereotactic brain injections, fiber photometry recordings, multi-unit in vivo electrophysiological recordings, optogenetic manipulation, and anterograde trans-synaptic tracing. Results: We analyzed mice performing standardized T-maze turning tasks and revealed that anatomically distinct RN neuronal ensembles exhibit direction-selective activity patterns. These neurons demonstrate preferential activation during ipsilateral turning movements, with activity onset consistently occurring after movement initiation. We establish a causal relationship between RN neuronal activity and directional motor control: selective activation of RN glutamatergic neurons facilitates ipsilateral turning, whereas temporally precise inhibition significantly impairs the execution of these movements. Anterograde trans-synaptic tracing using H129 reveals that RN neurons selectively project to spinal interneuron populations responsible for ipsilateral flexion and coordinated limb movements. Conclusions: These findings offer a framework for understanding asymmetric motor control in the brain. This work redefines the RN as a specialized hub within midbrain networks that mediate lateralized movements and offers new avenues for neuromodulatory treatments for neurodegenerative and post-injury motor disorders. Full article

To address overfitting due to limited sample size, and the challenges posed by “Spectral Homogeneity with Material Heterogeneity (SHMH)” and “Material Consistency with Spectral Divergence (MCSD)”—which arise from subtle spectral differences and limited classification accuracy—this study proposes a deep integration model that combines the Adaptive Boosting (AdaBoost) algorithm with a convolutional recurrent neural network (CRNN). The model adopts a dual-branch architecture integrating a 2D-CNN and gated recurrent unit to effectively fuse spatial and spectral features of rock samples, while the integration of the AdaBoost algorithm optimizes performance by enhancing system stability and generalization capability. The experiment used a hyperspectral dataset containing 81 rock samples (46 igneous rocks and 35 metamorphic rocks) and evaluated model performance through five-fold cross-validation. The results showed that the proposed 2D-CRNN-AdaBoost model achieved 92.55% overall accuracy, which was significantly better than that of other comparative models, demonstrating the effectiveness of multimodal feature fusion and ensemble learning strategy. Full article