Search Results (2,093)

Knowledge tracing, the computational modeling of student learning progression through sequential educational interactions, represents a critical component for adaptive learning systems and personalized education platforms. However, existing approaches face a fundamental trade-off between predictive accuracy and interpretability: deep sequence models excel at capturing complex temporal dependencies in student interaction data but lack transparency in their decision-making processes, while probabilistic graphical models provide interpretable causal relationships but struggle with the complexity of real-world educational sequences. We propose a hybrid architecture that integrates transformer-based sequence modeling with structured Bayesian causal networks to overcome this limitation. Our dual-pathway design employs a transformer encoder to capture complex temporal patterns in student interaction sequences, while a differentiable Bayesian network explicitly models prerequisite relationships between knowledge components. These pathways are unified through a cross-attention mechanism that enables bidirectional information flow between temporal representations and causal structures. We introduce a joint training objective that simultaneously optimizes sequence prediction accuracy and causal graph consistency, ensuring learned temporal patterns align with interpretable domain knowledge. The model undergoes pre-training on 3.2 million student–problem interactions from diverse MOOCs to establish foundational representations, followed by domain-specific fine-tuning. Comprehensive experiments across mathematics, computer science, and language learning demonstrate substantial improvements: 8.7% increase in AUC over state-of-the-art knowledge tracing models (0.847 vs. 0.779), 12.3% reduction in RMSE for performance prediction, and 89.2% accuracy in discovering expert-validated prerequisite relationships. The model achieves a 0.763 F1-score for early at-risk student identification, outperforming baselines by 15.4%. This work demonstrates that sophisticated temporal modeling and interpretable causal reasoning can be effectively unified for educational applications. Full article

Background/Objectives: Bones are essential components of the human body, providing structural support, enabling mobility, storing minerals, and protecting internal organs. Bone fractures (BFs) are common injuries that result from excessive physical force and can lead to serious complications, including bleeding, infection, impaired oxygenation, and long-term disability. Early and accurate identification of fractures through radiographic imaging is critical for effective treatment and improved patient outcomes. However, manual evaluation of X-rays is often time-consuming and prone to diagnostic errors due to human limitations. To address this, artificial intelligence (AI), particularly deep learning (DL), has emerged as a powerful tool for enhancing diagnostic precision in medical imaging. Methods: This research introduces a novel convolutional neural network (CNN) model, the Multi-Level Feature Fusion Network (MLFNet), designed to capture and integrate both low-level and high-level image features. The model was evaluated using the Bone Fracture Multi-Region X-ray (BFMRX) dataset. Preprocessing steps included image normalization, resizing, and contrast enhancement to ensure stable convergence, reduce sensitivity to lighting variations in radiographic images, and maintain consistency. Ablation studies were conducted to assess architectural variations, confirming the model’s robustness and generalizability across data distributions. MLFNet’s high accuracy, interpretability, and efficiency make it a promising solution for clinical deployment. Results: MLFNet achieved an impressive accuracy of 99.60% as a standalone model and 98.81% when integrated into hybrid ensemble architectures with five leading pre-trained DL models. Conclusions: The proposed approach supports timely and precise fracture detection, optimizing the diagnostic process and reducing healthcare costs. This approach offers significant potential to aid clinicians in fields such as orthopedics and radiology, contributing to more equitable and effective patient care. Full article

High energy-consuming industrial parks are characterized by high base-load-to-peak-valley ratios, overlapping production cycles, and megawatt-scale step changes, which significantly complicate short-term load forecasting. To tackle these challenges, this study proposes a novel forecasting framework that combines hierarchical multimodal decomposition with a hybrid deep learning architecture. First, Maximal Information Coefficient (MIC) analysis is applied to identify key input features and eliminate redundancy. The load series is then decomposed in two stages: seasonal-trend decomposition uses the Loess (STL) isolates trend and seasonal components, while variational mode decomposition (VMD) further disaggregates the residual into multi-scale modes. This hierarchical approach enhances signal clarity and preserves temporal structure. A parallel neural architecture is subsequently developed, integrating an Informer network to model long-term trends and a bidirectional gated recurrent unit (BiGRU) to capture short-term fluctuations. Case studies based on real-world load data from a typical industrial park in northeastern China demonstrate that the proposed model achieves significantly improved forecasting accuracy and robustness compared to benchmark methods. These results provide strong technical support for fine-grained load prediction and intelligent dispatch in high energy-consuming industrial scenarios. Full article

Traffic congestion remains a major urban challenge, impacting economic productivity, environmental sustainability, and commuter well-being. This systematic review investigates how artificial intelligence (AI) techniques contribute to detecting traffic congestion. Following the SPAR-4-SLR protocol, we analyzed 44 peer-reviewed studies covering three data categories—spatiotemporal, probe, and hybrid/multimodal—and four AI model types—shallow machine learning (SML), deep learning (DL), probabilistic reasoning (PR), and hybrid approaches. Each model category was evaluated against metrics such as accuracy, the F1-score, computational efficiency, and deployment feasibility. Our findings reveal that SML techniques, particularly decision trees combined with optical flow, are optimal for real-time, low-resource applications. CNN-based DL models excel in handling unstructured and variable environments, while hybrid models offer improved robustness through multimodal data fusion. Although PR methods are less common, they add value when integrated with other paradigms to address uncertainty. This review concludes that no single AI approach is universally the best; rather, model selection should be aligned with the data type, application context, and operational constraints. This study offers actionable guidance for researchers and practitioners aiming to build scalable, context-aware AI systems for intelligent traffic management. Full article

In this study, a lightweight deep neural network framework WeedLoc3D based on multi-task learning is proposed to meet the demand of accurate three-dimensional positioning of weed targets in automatic laser weeding. Based on a single RGB image, it both locates the 2D keypoints (growth points) of weeds and estimates the depth with high accuracy. This is a breakthrough from the traditional thinking. To improve the model performance, we introduce several innovative structural modules, including Gated Feature Fusion (GFF) for adaptive feature integration, Hybrid Domain Block (HDB) for dealing with high-frequency details, and Cross-Branch Attention (CBA) for promoting synergy among tasks. Experimental validation on field data sets confirms the effectiveness of our method. It significantly reduces the positioning error of 3D keypoints and achieves stable performance in diverse detection and estimation tasks. The demonstrated high accuracy and robustness highlight its potential for practical application. Full article

This review explores the evolution and current state of machine learning (ML) and artificial intelligence (AI) applications in direct energy deposition (DED) and wire arc additive manufacturing (WAAM) processes. A Python-based automated search script was developed to systematically retrieve relevant literature using the Crossref API, yielding around 370 papers published between 2010 and July 2025. The study identifies significant growth in ML-related DED research starting in 2020, with increasing adoption of advanced techniques such as deep learning, fuzzy logic, and hybrid physics-informed models. A year-by-year trend analysis is presented, and a comprehensive categorization of the literature is provided to highlight dominant application areas, including process optimization, real-time monitoring, defect detection, and melt pool prediction. Key challenges, such as limited closed-loop control, lack of generalization across systems, and insufficient modeling of deposition-location effects, are discussed. Finally, future research directions are outlined, emphasizing the need for integrated thermo-mechanical models, uncertainty quantification, and adaptive control strategies. This review serves as a resource for researchers aiming to advance intelligent control and predictive modeling in DED-based additive manufacturing. Full article

In the Ultra-High-Definition (UHD) domain, blind image quality assessment remains challenging due to the high dimensionality of UHD images, which exceeds the input capacity of deep learning networks. Motivated by the visual discrepancies observed between high- and low-quality images after down-sampling and Super-Resolution (SR) reconstruction, we propose a SUper-Resolved Pseudo References In Dual-branch Embedding (SURPRIDE) framework tailored for UHD image quality prediction. SURPRIDE employs one branch to capture intrinsic quality features from the original patch input and the other to encode comparative perceptual cues from the SR-reconstructed pseudo-reference. The fusion of the complementary representation, guided by a novel hybrid loss function, enhances the network’s ability to model both absolute and relational quality cues. Key components of the framework are optimized through extensive ablation studies. Experimental results demonstrate that the SURPRIDE framework achieves competitive performance on two UHD benchmarks (AIM 2024 Challenge, PLCC = 0.7755, SRCC = 0.8133, on the testing set; HRIQ, PLCC = 0.882, SRCC = 0.873). Meanwhile, its effectiveness is verified on high- and standard-definition image datasets across diverse resolutions. Future work may explore positional encoding, advanced representation learning, and adaptive multi-branch fusion to align model predictions with human perceptual judgment in real-world scenarios. Full article

Obesity is a major public health challenge linked to increased risks of chronic diseases. Effective prevention and intervention strategies require accurate classification and identification of key determinants. This study aims to develop a robust deep learning framework to enhance the accuracy and interpretability of obesity classification using comprehensive datasets, and to compare its performance with both traditional and state-of-the-art deep learning models. We propose a hybrid deep learning framework that combines an improved Mamba model with an attention-enhanced bidirectional LSTM (ABi-LSTM). The framework utilizes the Obesity and CDC datasets. A feature tokenizer is integrated into the Mamba model to improve scalability and representation learning. Channel-independent processing is employed to prevent overfitting through independent feature analysis. The ABi-LSTM component is used to capture complex temporal dependencies in the data, thereby enhancing classification performance. The proposed framework achieved an accuracy of 93.42%, surpassing existing methods such as ID3 (91.87%), J48 (89.98%), Naïve Bayes (90.31%), Bayesian Network (89.23%), as well as deep learning-based approaches such as VAE (92.12%) and LightCNN (92.50%). Additionally, the model improved sensitivity to 91.11% and specificity to 92.34%. The hybrid model demonstrates superior performance in obesity classification and determinant identification compared to both traditional and advanced deep learning methods. These results underscore the potential of deep learning in enabling data-driven personalized healthcare and targeted obesity interventions. Full article

Traditional vernacular architecture is an important component of historical and cultural heritage, and the accurate identification of its construction period is of great significance for architectural heritage conservation, historical research, and urban–rural planning. However, traditional methods for period identification are labor-intensive, potentially damaging to buildings, and lack sufficient accuracy. To address these issues, this study proposes a deep learning-based method for classifying the construction periods of traditional vernacular architecture. A dataset of traditional vernacular architecture images from the Longzhong region of Gansu Province was constructed, covering four periods: before 1911, 1912–1949, 1950–1980, and from 1981 to the present, with a total of 1181 images. Through comparative analysis of three mainstream models—ResNet50, EfficientNet-b4, and Vision Transformer—we found that EfficientNet demonstrated optimal performance in the classification task, achieving Accuracy, Precision, Recall, and F1-scores of 85.1%, 81.6%, 81.0%, and 81.1%, respectively. These metrics surpassed ResNet50 by 1.4%, 1.3%, 0.5%, and 1.2%, and outperformed Vision Transformer by 8.1%, 9.1%, 9.5%, and 9.1%, respectively. To further improve feature extraction and classification accuracy, we propose the “local–global feature joint learning network architecture” (DualBranchEfficientNet). This dual-branch design, comprising a global feature branch and a local feature branch, effectively integrates global structure with local details and significantly enhances classification performance. The proposed architecture achieved Accuracy, Precision, Recall, and F1-scores of 89.6%, 87.7%, 86.0%, and 86.7%, respectively, with DualBranchEfficientNet exhibiting a 2.0% higher Accuracy than DualBranchResNet. To address sample imbalance, a hybrid triplet loss function (Focal Loss + Triplet Loss) was introduced, and its effectiveness in identifying minority class samples was validated through ablation experiments. Experimental results show that the DualBranchEfficientNet model with the hybrid triplet loss outperforms traditional models across all evaluation metrics, particularly in the data-scarce 1950–1980 period, where Recall increased by 7.3% and F1-score by 4.1%. Finally, interpretability analysis via Grad-CAM heat maps demonstrates that the DualBranchEfficientNet model incorporating hybrid triplet loss accurately pinpoints the key discriminative regions of traditional dwellings across different eras, and its focus closely aligns with those identified by conventional methods. This study provides an efficient, accurate, and scalable deep learning solution for the period identification of traditional vernacular architecture. Full article

Wetlands are an important part of coastal ecosystems but are under increasing pressure from climate change-induced sea-level rise and flooding, in addition to development pressures associated with increasing human populations. The change in tidal events and their intensity due to sea-level rise is also reshaping and challenging the vitality of existing wetland systems, requiring more intensive localized studies to identify future-focused restoration and conservation strategies. To support this endeavor, this study utilizes tide gauge datasets from the Australian Bureau of Meteorology (BOM) for maximum sea-level (Hmax) prediction and Landsat Collection surface reflectance datasets obtained from the United States Geological Survey (USGS) database to detect and project patterns of change in the Maroochy River floodplain of Queensland, Australia. This study developed an efficient hybrid deep learning model combining a Convolutional Neural Network and Bidirectional Long Short-Term Memory (CNNBiLSTM) architecture for the prediction of maximum sea-level and tidal events. The proposed model significantly outperformed three benchmark models (Multiple Linear Regression (MLR), Support Vector Regression (SVR), and CatBoost) in achieving a high correlation coefficient (r = 0.9748) for maximum sea-level prediction. To further address the increasing frequency and intensity of tidal events linked to sea-level rise, a CNNBiLSTM classification model was also developed, achieving 96.72% accuracy in predicting extreme tidal occurrences. This study identified a significant positive linear increase in sea-level rise of 0.016 m/year between 2014 and 2024. Wetland change detection using Landsat imagery along the Maroochy River floodplain also identified a substantial vegetation loss of 395.64 hectares from 2009 to 2023. These findings highlight the strong potential of integrating deep learning and remote sensing for improved prediction and assessment of sea-level extremes and coastal ecosystem changes. The study outcomes provide valuable insights for informing not only conservation and restoration activities but also for providing localized projections of future change necessary for the progression of effective climate adaptation and mitigation strategies. Full article

Terahertz (THz) communication is a key technology for sixth-generation (6G) networks, offering ultra-high data rates, low latency, and massive connectivity. However, the THz band faces significant propagation challenges, including high path loss, molecular absorption, and susceptibility to blockage. Reconfigurable intelligent surfaces (RISs) have emerged as an effective solution to overcome these limitations by reconfiguring the wireless environment through passive beam steering. In this work, we propose a novel framework, namely the optimized deep reinforcement learning transformer (ODRL-Transformer), to maximize the sum rate in RIS-assisted THz systems. The framework integrates a Transformer encoder for extracting temporal and contextual features from sequential channel observations, a DRL agent for adaptive beamforming and phase shift control, and a hybrid biogeography-based optimization (HBBO) algorithm for tuning the hyperparameters of both modules. This design enables efficient long-term decisionmaking and improved convergence. Extensive simulations of dynamic THz channel models demonstrate that ODRL-Transformer outperforms other optimization baselines in terms of the sum rate, convergence speed, stability, and generalization. The proposed model achieved an error rate of 0.03, strong robustness, and fast convergence, highlighting its potential for intelligent resource allocation in next-generation RIS-assisted THz networks. Full article

The rise of streaming services and online story-sharing has led to a vast amount of cinema and television content being viewed and reviewed daily by a worldwide audience. It is a unique challenge to grasp the nuanced insights of these reviews, particularly as context, emotion, and specific components like acting, direction, and storyline intertwine extensively. The aim of this study is to address said complexity with a new hybrid Multi Criteria Decision-Making MCDM model that combines the Deck of Cards Method (DoCM) with the Circular Pythagorean Fuzzy Set (CPFS) framework, retaining the symmetry of information. The study is conducted on a simulated dataset to demonstrate the framework and outline the plan for approaching real-world press reviews. We postulate a more informed mechanism of assessing and choosing the most appropriate deep learning assembler, such as the transformer version, the hybrid Convolutional Neural Network CNN-RNN, and the attention-based framework of aspect-based sentiment mapping in film and television reviews. The model leverages both the cognitive ease of the DoCM and the expressive ability of the Pythagorean fuzzy set (PFS) in a circular relationship setting possessing symmetry, and can be applied to various decision-making situations other than the interpretation of media sentiments. This enables decision-makers to intuitively and flexibly compare alternatives based on many sentiment-relevant aspects, including classification accuracy, interpretability, computational efficiency, and generalization. The experiments are based on a hypothetical representation of media review datasets and test whether the model can combine human insight with algorithmic precision. Ultimately, this study presents a sound, structurally clear, and expandable framework of decision support to academicians and industry professionals involved in converging deep learning and opinion mining in entertainment analytics. Full article

Down syndrome (DS) is one of the prevalent chromosomal disorders, representing distinctive craniofacial features and a range of developmental and medical challenges. Due to the lack of clinical expertise and high infrastructure costs, access to genetic testing is restricted to resource-constrained clinical settings. There is a demand for developing a non-invasive and equitable DS screening tool, facilitating DS diagnosis for a wide range of populations. In this study, we develop and validate a robust, interpretable deep learning model for the early detection of DS using facial images of infants. A hybrid feature extraction architecture combining RegNet X–MobileNet V3 and vision transformer (ViT)-Linformer is developed for effective feature representation. We use an adaptive attention-based feature fusion to enhance the proposed model’s focus on diagnostically relevant facial regions. Bayesian optimization with hyperband (BOHB) fine-tuned extremely randomized trees (ExtraTrees) is employed to classify the features. To ensure the model’s generalizability, stratified five-fold cross-validation is performed. Compared to the recent DS classification approaches, the proposed model demonstrates outstanding performance, achieving an accuracy of 99.10%, precision of 98.80%, recall of 98.87%, F1-score of 98.83%, and specificity of 98.81%, on the unseen data. The findings underscore the strengths of the proposed model as a reliable screening tool to identify DS in the early stages using the facial images. This study paves the foundation to build equitable, scalable, and trustworthy digital solution for effective pediatric care across the globe. Full article

Accurate state-of-charge (SOC) estimation is crucial in smart-building energy management systems, where rooftop photovoltaics and lithium-ion energy storage systems must be coordinated to align renewable generation with real-time demand. This paper introduces a novel, modular hybrid framework for SOC estimation, which synergistically combines the predictive power of artificial neural networks (ANNs), the logical consistency of finite state automata (FSA), and an adaptive dynamic supervisor layer. Three distinct ANN architectures—feedforward neural network (FFNN), long short-term memory (LSTM), and 1D convolutional neural network (1D-CNN)—are employed to extract comprehensive temporal and spatial features from raw data. The inherent challenge of ANNs producing physically irrational SOC values is handled by processing their raw predictions through an FSA module, which constrains physical validity by applying feasible transitions and domain constraints based on battery operational states. To further enhance the adaptability and robustness of the framework, two advanced supervisor mechanisms are developed for model selection during estimation. A lightweight rule-based supervisor picks a model transparently using recent performance scores and quick signal heuristics, whereas a more advanced double deep Q-network (DQN) reinforcement-learning supervisor continuously learns from reward feedback to adaptively choose the model that minimizes SOC error under changing conditions. This RL agent dynamically selects the most suitable ANN+FSA model, significantly improving performance under varying and unpredictable operational conditions. Comprehensive experimental validation demonstrates that the hybrid approach consistently outperforms raw ANN predictions and conventional extended Kalman filter (EKF)-based methods. Notably, the RL-based supervisor exhibits good adaptability and achieves lower error results in challenging high-variance scenarios. Full article

Mining subsidence is a pervasive geohazard in coal basins, and precise and reliable deformation monitoring is essential to effective risk mitigation. Conventional time-series Interferometric Synthetic Aperture Radar (InSAR) suffers from vegetation-induced decorrelation and atmospheric delays. Most predictive models leverage only temporal information. We introduced an integrated DS InSAR + CNN LSTM framework for subsidence monitoring and forecasting. Forty-three Sentinel-1A scenes (2017–2018), corrected with Generic Atmospheric Correction Online Service for InSAR (GACOS) data, were processed to derive cumulative deformation, cross-validated against multi-view SBAS InSAR, and used to train a CNN LSTM network that predicts trends one year in advance. The findings indicate that (1) DS InSAR provides 2.83 times the monitoring density of SBAS InSAR, with deformation rate R² = 0.83, RMSE = 0.0028 m/a, and MAE = 0.0019 m/a at common pixels. The RMS average decrease in GACOS atmospheric delay phase correction is 2.52 mm. (2) High- and low-settlement zones comprise 0.11% and 92.32% of the area, respectively; maximum velocity reaches 190.61 mm/a, with a cumulative subsidence of −338.33 mm. (3) Across the five zones with the most severe subsidence, the CNN–LSTM model attains R² values of 0.97–0.99 and RMSE below 1 mm, markedly outperforming the standalone LSTM network. (4) Deformation correlated strongly with geological structures, groundwater decline (R² = 0.66–0.78), and precipitation (slope > 0.33), highlighting coupled natural and anthropogenic control. Full article