Search Results (994)

Laser powder bed fusion (LPBF) can fabricate intricate metal components but is prone to defects, such as porosity and cracks, that degrade performance. We present an in situ monitoring framework that fuses structure-borne acoustic emission (AE) and coaxial two-color pyrometry acquired synchronously at 1 MHz. Modality-specific encoders are pretrained separately, their latent representations are exported, and a lightweight feature-level fusion classifier with two binary heads predicts crack-like and porosity-like indications. Evaluation uses a held-out grouped experiment/build-machine-part split with independent Archimedes density and micro-CT ground truth. On the held-out test set, the fused model achieved F1 = 0.974 for crack-like detection and F1 = 0.987 for porosity-like detection, with AUROC = 0.998 and 0.993, respectively. Recall was 1.00 for both heads, corresponding to false-positive rates of 11.18% for crack-like and 0.945% for porosity-like indications. These results support synchronized AE-pyrometry fusion as a promising high-sensitivity in situ screening approach for LPBF. A later matched within-framework ablation campaign was also performed under stricter checkpoint-screening rules to compare AE + PY + Aux, AE + PY, AE-only, and PY-only variants under a common grouped-split protocol. Together, these results support multimodal monitoring while highlighting the need for explicit coupon/geometry-stratified reporting and for separately architecture-optimized unimodal baselines. Full article

Serum contains diverse proteins whose concentrations vary with pathological conditions such as cancer, liver disease, neurological disorder, and infections. Conventional methods like serum protein electrophoresis (SPEP) and enzyme-linked immunosorbent assay (ELISA) are gold standards for protein identification; however, they are time-consuming and can miss abnormal serum protein levels. Inspired by chemical nose sensing based on selective sensor–analyte interactions, we synthesized five pyrene-conjugated fluorescent polymers (PFPs) with distinct side-chain head groups to construct a multichannel fluorescence sensor array. These polymers were screened for sensitivity to changes in serum protein levels using linear discriminant analysis (LDA), a machine learning method. This process led to the successful discovery of two PFPs that effectively detect protein level fluctuations. These PFPs provided a sensitive sensor array capable of generating a high-content response pattern (fingerprint) with six fluorescence channels. This sensor array successfully discriminated protein level fluctuations in serum with 98% jackknife classification accuracy and 95% unknown identification accuracy. This polymer sensor array holds strong potential as a diagnostic tool for serum-based samples and can be extended to other applications related to protein identification. Full article

Artificial intelligence (AI) is rapidly reshaping head and neck surgical oncology by augmenting decision-making across the full perioperative continuum. This state-of-the-art review aims to provide head and neck surgical oncologists with a conceptual framework for understanding and critically appraising AI tools entering clinical practice, summarizing how machine learning, deep learning, and generative AI are being integrated into contemporary surgical workflows. Preoperative applications include detection of occult nodal metastasis and extranodal extension. Intraoperative innovations include augmented reality-assisted navigation, real-time margin assessment, and improving visual clarity and tissue handling for robotic platforms. Postoperatively, AI can predict complications like free flap failure and oncologic outcomes. Large language models are being operationalized for clinician-facing applications such as documentation and inbox support, as well as patient-facing education. Despite promising results, broad clinical deployment remains limited by concerns about privacy, validation, reliability, safety, and ethics. Widespread adoption will require prospective clinical trials, robust governance, and human-centered workflows that ensure AI remains a safe, assistive copilot. Full article

To address collision risks in teleoperated mobile robotic vehicles, this study proposes a Human–Machine Interaction-based Advanced Smooth Shared Control (ASSC) system aimed at enhancing obstacle avoidance and achieving smooth shared control between human operators and the automation system. The ASSC system integrates a novel approach using predictive vectors to represent the vehicle’s heading position, automatically adjusting the steering position upon obstacle detection to ensure smooth collision avoidance without changing the driver’s perception. Feedback forces applied to the steering wheel are calculated through an artificial potential field algorithm. Twenty participants were invited to operate the vehicle, providing feedback on the ASSC system’s performance relative to conventional obstacle avoidance methods. Performance metrics such as the effects of communication delays, Time to Complete the Task (TTC), ASSC effectiveness, performance of the delay impact on the ASSC system, and the Number of Obstacle Collisions (NOC) are analyzed. The results demonstrate that the ASSC system significantly outperforms traditional obstacle avoidance methods, providing more precise control in teleoperation. Statistical analysis indicates that the ASSC system improves safety, comfort and operational performance by 12.8%. This research highlights the ASSC system as a promising solution for enhancing automation, safety, and human–machine interaction in teleoperated mobile robotic vehicles. Full article

Credit card fraud detection remains a critical challenge in financial security, characterized by severe class imbalance and the need to capture complex temporal patterns in transaction sequences. Traditional machine learning approaches treat transactions as independent events, failing to model the sequential nature of user behavior and suffering from inadequate handling of minority class samples. In this paper, we propose an integrated framework that combines generative modeling and time-aware sequential learning for credit card fraud detection. Our approach addresses two fundamental limitations: (1) we model transaction histories as temporal sequences using a Transformer-based architecture that captures both long-term dependencies and abrupt behavioral changes through multi-head self-attention mechanisms, and (2) we employ CTGAN to generate high-quality synthetic fraudulent samples, providing more effective oversampling than conventional techniques like SMOTE. The Time-Aware Transformer incorporates temporal encoding and position-aware attention to preserve transaction order and time intervals, while CTGAN learns the complex conditional distributions of fraudulent transactions to produce realistic synthetic samples. We evaluate our framework on the IEEE-CIS Fraud Detection dataset, demonstrating significant improvements over representative classical and sequential deep-learning baselines. Experimental results show that our method achieves superior performance with an AUC-ROC of 0.982, precision of 0.891, recall of 0.876, and F1-score of 0.883, outperforming the representative baselines considered in this study, including traditional machine learning models, standalone deep learning architectures, and supervised sequential neural models. Ablation studies confirm the individual contributions of both the sequential modeling component and the generative oversampling strategy. Our work demonstrates that combining temporal sequence modeling with generative synthesis provides a robust solution for imbalanced fraud detection, with potential applications extending to other domains requiring sequential pattern recognition under extreme class imbalance. Full article

Retrieving farmland soil water content with both high accuracy and physical interpretability remains a significant challenge, particularly for winter wheat. To bridge the gap between purely empirical data-driven approaches and mechanistic scattering models, this study proposed a novel hybrid framework that integrates an improved water cloud model (IWCM) with machine learning algorithms. Multi-modal unmanned aerial vehicle (UAV) experiments were conducted during the heading stage of winter wheat over two consecutive years (2024–2025) using a synchronized system equipped with a miniature synthetic aperture radar (MiniSAR) and a multi-spectral sensor. The core innovation of the proposed framework lies in the IWCM, which explicitly decouples vegetation and soil scattering contributions by incorporating fractional vegetation cover, thereby deriving physically meaningful soil backscatter coefficients from complex microwave signals. Unlike traditional methods that treat remote sensing variables as black box inputs, our approach employed these physics-derived features to guide data-driven modeling. Four feature input schemes including spectral reflectance, vegetation indices, MiniSAR polarimetric parameters, and their multi-source fusion were systematically evaluated using back propagation neural network (BPNN) and random forest (RF) regressors. The results demonstrated that the proposed framework significantly enhances retrieval performance. Notably, the RF model driven by spectral band reflectance within this physically constrained architecture achieved optimal accuracy, with a coefficient of determination (R²) of 0.865, a mean absolute error (MAE) of 0.0152, and a root mean square error (RMSE) of 0.0197. Compared to purely empirical approaches, the IWCM significantly improved the physical interpretability of microwave polarimetric characteristics, enabling the multi-source data fusion to better represent the interactions among vegetation, soil, and microwave scattering. This study demonstrated that integrating mechanistic models with multi-source UAV remote sensing data not only improves soil water content retrieval accuracy in winter wheat fields but also provides a valuable reference for developing operationally applicable and physically interpretable farmland soil water content monitoring systems. Full article

Financial time series forecasting remains a challenging task due to the high non-stationarity and concept drift inherent to market data. Existing deep learning models, such as LSTMs and transformers, typically employ static weights after training, limiting their ability to adapt to rapid market regime shifts (e.g., from trends to reversals). To bridge this gap between static parameters and dynamic environments, we propose a novel framework named Game-Theoretic HyperNetwork (GTH-Net), which introduces a context-aware meta-learning mechanism to achieve adaptive forecasting. Specifically, we first introduce an Evolutionary Game-Theoretic Correction Module (E-GTCM) to explicitly extract latent buying and selling pressure based on market microstructure priors through an iterative gated evolution process. Subsequently, we propose a HyperNetwork-based fusion mechanism that treats the extracted game state as a meta-context to dynamically generate the weights of the forecasting head. This allows the model to automatically switch its prediction rules in response to shifting market regimes. Extensive experiments on real-world stock datasets demonstrate that GTH-Net significantly outperforms baselines in terms of machine learning predictive accuracy and simulated financial profitability. Furthermore, ablation studies and parameter analysis confirm that the dynamic weight generation mechanism effectively captures market reversals caused by overcrowded trades. Full article

With the rapid development in data mining and learning analytics, integrating predictive analytics into educational data has become increasingly critical for supporting students’ learning trajectories. In many schooling systems, the academic tracks (such as Liberal Arts or Science) and the performance of junior high school students can substantially shape their subsequent university pathways and career planning. Despite the long-term impact of these decisions, academic track selections and the evaluation of students’ potential are often made without systematic and evidence-based guidance. Predictive computer applications can assist, but the training of accurate models and the selection of adequate features remain key challenges. This paper details our process of engineering such an application comprising two tasks based on 1357 real-world junior high school academic performance records. The first task applies a classification approach to predict students’ academic track orientation, while the second task employs a multi-output regression model to forecast students’ future academic performance in senior high school. Our approach shows that the stacking ensemble model achieved a classification accuracy of 85.76%, whereas the Bi-LSTM model with multi-head attention attained an overall R² exceeding 82% in performance forecasting; both models demonstrated strong and reliable predictive capability. Moreover, the proposed approach provides inherent interpretability by decomposing predictions at the subject level. Feature importance analysis reveals how different academic subjects contribute variably to both academic track decisions and future academic performance, offering actionable insights for academic counselling and future study planning. By bridging predictive modelling with students’ educational and career planning needs, this study advances the practical application of educational data mining and provides support for evidence-based academic guidance and future career choices in real-world contexts. Full article

To address traditional beer detection limitations, this study proposes a rapid NIRS-based method for detecting key indicators and verifying authenticity. Designing Single-task (STL) and Multi-task learning (MTL) strategies, it employs Variable Importance in Projection for wavelength selection. Deep spectral features were extracted utilizing a Multi-Head Attention (MHA)-fused Convolutional Neural Network (CNN-MHA), Long Short-Term Memory (LSTM-MHA), and hybrid CNN-LSTM-MHA networks. To further enhance model performance, the Bayesian Optimization Algorithm globally optimized network hyperparameters in STL, alongside hyperparameters and multi-task loss weights in MTL. Partial least squares regression, support vector machine regression, and partial least squares discriminant analysis models were established using these features. Results indicate that the MTL-based CNN-LSTM-MHA network effectively learns shared features across multiple tasks, significantly improving model generalization. Specifically, the coefficients of determination ($R^2$) for alcohol content and original wort concentration in the validation set were 0.996 and 0.997, respectively, with relative root mean square errors (rRMSE) of 2.024% and 2.515%. In the independent test set, the $R^2$ were 0.995 and 0.991, with rRMSE of 2.515% and 2.087%, respectively. Furthermore, 100% classification accuracy was achieved across all datasets. This method provides an efficient technical solution for beer market regulation and real-time detection in production processes. Full article

Background/Objectives: Radiomics-based machine learning models have demonstrated high accuracy in differentiating benign from malignant orbital masses, with early studies suggesting performance comparable to expert radiologists. However, translation into clinical practice remains limited due to dataset constraints, including retrospective study designs, single-center cohorts, and underrepresentation of diverse patient populations. This review aims to evaluate the current evidence supporting radiomics in orbital disease while critically examining barriers to generalizability and equity across ophthalmology, otolaryngology, and plastic surgery. Methods: A narrative literature review was conducted to assess radiomics applications in orbital oncology and reconstruction. Studies evaluating diagnostic accuracy, margin assessment, postoperative surveillance, and surgical planning across ophthalmology, head and neck surgery, and reconstructive surgery were analyzed, with particular attention paid to dataset composition, validation strategies, and imaging standardization. Results: Radiomics models demonstrated high diagnostic performance in differentiating orbital tumors, optimizing surgical planning, and aiding postoperative monitoring. However, most studies relied on small, homogeneous datasets lacking racial, ethnic, and pediatric representation. External validation was uncommon, and imaging heterogeneity limited reproducibility. These deficiencies restrict the clinical translation of radiomics and risk exacerbating healthcare disparities, particularly among underrepresented populations. Conclusions: Radiomics holds promise as a precision medicine tool for orbital diagnosis, surgical navigation, and postoperative care. Nevertheless, its clinical adoption is constrained by dataset bias, lack of standardization, and limited prospective validation. Future progress requires multi-institutional, demographically diverse datasets and standardized imaging protocols to ensure equitable and generalizable implementation across specialties. Full article

Wireless sensor networks (WSNs) are vulnerable to physical attacks in which adversaries gain partial or full control of sensor nodes, compromising the integrity of the network. Conventional security mechanisms impose excessive computational overhead and are not well suited to resource-constrained WSN devices. This paper proposes a lightweight, two-layer intrusion detection system (IDS) that integrates blockchain (BC) technology with machine learning for physical attack detection in WSNs. The first layer employs a lightweight BC protocol among cluster heads (CHs) and the base station (BS) to detect data integrity violations through hash-based consensus. The second layer applies an artificial neural network (ANN) at the base station to detect attacks that bypass blockchain verification, without imposing any processing load on sensor nodes. Simulation experiments on a 100-node WSN demonstrate that the combined system achieves 97.42% accuracy and 98.35% recall, outperforming five established classifiers and both standalone components. The system sustains detection rates above 99.98% under 30 simultaneous attackers and maintains reliable operation under packet loss conditions up to 10%. Full article

As mineral exploration increasingly targets deeper and more geologically complex terrains, the need for reliable predictive models becomes critical to mitigating exploration risk and improving cost efficiency. Correspondingly, the effectiveness of deep mineral exploration strategies depends substantially on the effectiveness and precision of three-dimensional mineral prospectivity mapping (3D MPM) models. However, the inherent spatial non-stationarity—where ore grade variability changes across geological domains—and the strongly skewed distribution of high-grade samples present a dual challenge. Conventional methods, which primarily rely on mean-based regression, often struggle to adequately address this dual challenge, limiting their predictive performance in complex geological settings. To address these issues, this paper proposes a pinball-loss-guided, global–local fusion Transformer model within a unified framework for 3D MPM. It leverages a multi-head self-attention mechanism with global–local fusion to capture long-range dependencies and global geological contexts, while incorporating local feature extraction modules to adaptively model spatially varying mineralization controls, jointly optimized through a pinball loss function to address mineralization distribution skewness. The proposed framework was first rigorously evaluated using the Xiadian gold deposit as a case study. Bootstrap analysis of the ablation experiments confirmed its predictive performance in terms of quantile-specific accuracy and prediction interval (PI) calibration. Ten rounds of random data splits provided further confirmation of the model’s stability. Subsequently, the validated model was applied to prospectivity mapping in unexplored regions, leading to the delineation of several high-potential exploration targets. Finally, comparative analyses with state-of-the-art machine learning methods were conducted, which further validated the competitive fitting capability of the proposed framework. Full article

Unbalanced supply voltage (USV) represents a critical power quality challenge in industrial environments, significantly degrading the performance, efficiency, and operational lifespan of three-phase induction motors. Accurate real-time estimation of sequence impedances ($Z_a,Z_b,Z_c$) and detection of the Negative Voltage Factor (NVF) are essential for effective condition monitoring and preventive maintenance strategies. While existing machine learning methods have demonstrated promising accuracy, they often rely on manual feature engineering, lack hierarchical representation learning, and treat impedance estimation and fault detection as isolated tasks. This paper proposes a unified Deep Multi-Task Learning framework that leverages Residual Multilayer Perceptron (ResMLP) architectures for feature-based learning and Temporal Convolutional Networks (TCNs) for end-to-end raw signal learning. Our contributions include: (1) introduction of a Multi-Head ResMLP architecture that jointly optimizes phase impedance and fault detection, achieving superior NVF accuracy (MAE = 0.0007) and a fault detection F1-score of 0.8831; (2) investigation of raw-voltage TCN models for voltage-only diagnostics, with analysis of the trade-offs between end-to-end learning and feature-based approaches; (3) extensive ablation studies demonstrating the impact of network depth, data augmentation, and training protocols on model generalization; and (4) deployment of PyTorch (v2.0.1)-based models suitable for embedded systems with real-time inference capabilities (2.3 ms per prediction). Experimental validation on a 1.1 kW three-phase motor dataset under diverse load conditions (0–10 Nm) and USV magnitudes (5–15 V) confirms the robustness and practical applicability of the proposed approach for industrial fault diagnosis and condition monitoring systems. Full article

Rapid non-destructive evaluation of the moisture content of freshly harvested paddy rice in the field is essential for determining the optimal harvest timing, ensuring high-quality rice production and energy savings. This study developed a non-destructive prediction model for the moisture content of paddy rice using near-infrared (NIR) spectroscopy combined with machine learning and deep learning techniques. Rice samples were collected weekly during the ripening period after heading, and NIR reflectance spectra were acquired in the range of 950–2200 nm. Seven spectral preprocessing techniques were applied; and the prediction models developed, using partial least squares regression, support vector regression, deep neural network, and one-dimensional convolutional neural networks (1D-CNNs) based on VGGNet and EfficientNet architectures. Among these, the EfficientNet-based 1D-CNN combined with Savitzky–Golay 1st order derivative preprocessing showed the highest performance, achieving an ${\mathrm{R}}_{\mathrm{p}}^2$ of 0.999 and an RMSE_P of 0.001 (Friedman test, p < 0.001; Kendall’s W = 0.97), significantly outperforming previous traditional machine learning models. The results demonstrate that the proposed prediction model enables highly accurate estimation of moisture content in freshly harvested paddy rice without requiring drying or milling. The proposed approach can be implemented across various agricultural operations, enabling optimal harvest timing, quality control during storage, energy efficient drying, and real-time monitoring via on-combine sensor systems. Full article

Motion sickness (MS) is a prevalent condition that can significantly degrade user comfort and immersion, particularly in virtual reality (VR) environments. Accurate prediction models are essential for early detection and mitigation of MS symptoms, thereby improving the overall VR experience. Most existing approaches rely on bio-physiological data acquired through body-mounted sensors, which may restrict user mobility and diminish immersion. This study proposes a less intrusive alternative, leveraging head and torso kinematic data for MS prediction. We introduce a hybrid Convolutional–Recurrent Neural Network (C-RNN) designed to capture both spatial and temporal features for enhanced classification accuracy. Using a dataset of 40 participants, the proposed C-RNN outperformed traditional machine learning models—including Support Vector Machines (SVMs), k-Nearest Neighbors (KNN), Decision Trees (DT), and a baseline Recurrent Neural Network (RNN)—across multiple evaluation metrics. The C-RNN achieved 85.63% accuracy, surpassing SVM (60%), KNN (73.75%), DT (74.38%), and RNN (81.88%), with corresponding gains in precision, recall, F1-score, and ROC AUC. These results demonstrate that head–torso motion patterns provide sufficient predictive signal for accurate MS detection, offering a non-intrusive, efficient alternative to physiological sensing that supports improved comfort and sustained immersion in VR. Full article