Search Results (100)

Entity-aware quality assessment may reduce false interpretations of electronic health record (EHR) data, but evidence from small, rule-aligned benchmarks cannot establish operational effectiveness. We revised H-StreamQ as a proof-of-concept framework and evaluated its laboratory component using the complete MIMIC-IV v3.1 labevents file (158,374,764 events; 313,442 patients). Ten thousand patients were sampled across laboratory-activity quintiles and split at patient level into training (6000), threshold-calibration (2000), and test (2000) groups. The independent test set contained 918,651 numeric laboratory events. Without excluding naturally alerted records, 54,788 mutually exclusive defects were introduced using subtle value shifts, unit/scale errors, mapping errors, delayed records, and patient-clustered correlated defects. Rules, a context-aware Isolation Forest, their union (Hybrid), a context-free Isolation Forest, Local Outlier Factor (LOF), and linear and radial-basis-function (RBF) One-Class support vector machines (OCSVMs) were compared at a threshold fixed by a 2.5% calibration alert budget. Patient-cluster bootstrap intervals and event-micro and patient-macro results were reported. Rules alone achieved the highest event-micro F1-score (0.637; 95% confidence interval [CI] 0.547–0.722), followed by Hybrid (0.576; 0.484–0.668) and RBF One-Class SVM (0.559; 0.433–0.670). Hybrid increased recall over rules by only 0.004 (95% CI 0.003–0.006) while reducing F1 by 0.061 and increasing the background-alert rate by 0.015. Context conditioning did not improve aggregate Isolation Forest performance. In six batch-level drift simulations, an exponentially weighted moving average (EWMA) and a fixed-window monitor detected 97–100% and 98–100% of changes, respectively, whereas a custom Hoeffding adaptive-window detector was more conservative and often missed smaller or recurrent changes. These results support H-StreamQ as an explainable research framework, not as a validated clinical or production system. Patient-macro F1, which weights every patient equally, was substantially lower than event-micro F1 for every method (rules 0.395 versus 0.637; Hybrid 0.320 versus 0.576), indicating that event-level performance is weighted towards high-activity patients. Precision and F1 are computed relative to injected synthetic labels and are not clinically adjudicated estimates. The entity-aware architecture spans patients, admissions, diagnoses, transfers, and dictionaries, but the quantitative detection benchmark evaluates the numeric laboratory component only; other entities are used for linkage and contextual attachment and are audited descriptively rather than evaluated against labels. Full article

One of the reasons for food waste is the inefficient use of by-products. This paper presents a current overview of technological solutions for processing waste, aimed at giving it a proverbial “second life.” To this end, the most important extraction, conversion, and valorization methods are summarized and compared. Extraction using green solvents (ionic liquids, deep eutectic solvents, supercritical fluids) and green methods such as micro-wave-assisted extraction (MAE) and ultrasound-assisted extraction (UAE) are discussed. Advances in biorecycling and biotechnological methods, taking into account the role of genetically modified microorganisms (GMOs), are also described. Attention was paid to innovative approaches and current trends, such as 3D food waste printing, hybrid technologies, and the concept of smart factories involving artificial intelligence (AI), machine learning (ML), and the Internet of Things (IoT). Although all of the mentioned food waste upcycling technologies are already in use, several hurdles limit their use to their full potential. A major problem is the smooth flow of food waste to upcycling points to ensure the required quality and processing optimization. Furthermore, societal and regulatory concerns related to the use of food waste for producing value-added dietary products need to be considered. Full article

The advanced manufacturing of metal micro-holes is of great significance in various fields of industrial production, including aerospace, automotive, electronics, and healthcare. New technologies are constantly emerging, including various multi-energy field manufacturing technologies, and the knowledge system is complex and intricate. The present article summarizes recent advancements in metal micro-hole manufacturing technologies, drawing parallels with existing laser processing and electrochemical processing technologies. The present systematic review has been conducted with the objective of providing a comprehensive overview of the latest methodologies. The present review paper is of particular significance in that it encompasses not only the fundamental principles and innovative process methods, but also the most recent research progress and current problems. Furthermore, a synopsis of the developmental trajectory of advanced sustainable manufacturing technology for micro-holes was furnished. Full article

Flexible capacitive sensors for electronic skins and soft robotic systems are expected to provide not only high-pressure sensitivity but also multifunctional sensing capabilities. However, conventional dielectric layer designs often suffer from a trade-off among multiple functionalities. To address this challenge, we developed a flexible sensor featuring a hybrid microstructured functional layer for pressure sensing, distance monitoring, and material identification. The functional layer was a polydimethylsiloxane (PDMS) film embedded with micro-sized sugar particles and patterned with microstructures on its surface. The pressure-sensing performance, such as pressing sensitivity, response time, and hysteresis, was first evaluated. The pressure sensitivity reached 3.55 × 10⁻² kPa⁻¹ at an applied force of 1 N, which is significantly greater than that of the sensor using either a flat PDMS layer or a PDMS film embedded solely with sugar particles. The hybrid microstructured sensor also exhibited superior performance in terms of hysteresis and repeatability. Moreover, the sensor was shown to measure the distance to an object with a sensitivity of 0.023 mm⁻¹. Furthermore, the robust identification of materials with different permittivities was demonstrated using the flexible sensor. Given its multifunctional, non-contact, and high-sensitivity capabilities, this flexible sensor holds significant potential for integration into advanced electronic skins, intelligent soft robotics for tactile object sorting, and human–-machine interfaces. Full article

Improving software quality remains a major challenge in modern software engineering. Among the different techniques used to address this issue, code smell detection plays an important role, as code smells may negatively affect software maintainability, readability, and evolution. Recent advances in machine learning, particularly pre-trained code representation models, have opened new perspectives for automating this task. In this study, code smell detection is formulated as a multi-label classification problem in order to reflect realistic scenarios where multiple smells may co-occur within the same code fragment. To enable the evaluation of pre-trained code representation models in a multi-label setting, six multi-label datasets containing source code were derived from two existing datasets, namely MLCQ and Fontana, by merging instances sharing common code smells. The study investigates multiple experimental configurations based on CodeBERT, GraphCodeBERT, and CodeT5, including direct transformer-based classification, sliding window processing for long code sequences, and hybrid approaches combining transformer embeddings with classical machine learning classifiers. Among the evaluated configurations, the hybrid approach based on CodeBERT embeddings combined with a Random Forest classifier achieved the best overall performance, reaching an accuracy of 0.91 and a micro-F1 score of 0.90 on the Fontana datasets. Full article

Additive–subtractive hybrid manufacturing (ASHM) allows for the rapid manufacturing of metal components with complex and precise geometries for ready-to-use or near-ready-to-use applications. Laser wire-directed energy deposition (LW-DED) can be used to quickly manufacture metal components, while CNC machining can achieve precise geometric tolerances. In this study, Ti-6Al-4V alloy specimens were fabricated using an LW-DED process combined with CNC machining and tested to evaluate the effects of ASHM on mechanical performance. Post fabrication, the Ti-6Al-4V material was evaluated through hardness mapping, monotonic tensile testing, and fully reversed axial fatigue testing. Vicker’s micro-hardness mapping showed a range of hardness results from 300 to 350 HV in the ASHM Ti-6Al-4V that remained consistent throughout the build. Tensile results showed a similar response to cast and wrought Ti-6Al-4V, with an average yield stress of 819.4 MPa, ultimate tensile strength of 935.5 MPa, and modulus of 119 GPa. When tested in fatigue, the material had a reduced life compared to wrought Ti-6Al-4V, which is attributed to defects originating from the additive process. While no run-outs were observed from the testing, the fatigue results remain aligned with trends reported for other methods of additively manufactured Ti-6Al-4V. Fully reversed high-cycle fatigue loading revealed that the ASHM-fabricated Ti-6Al-4V fell into a Basquin power-law fit with a fatigue strength coefficient of 1942 MPa with a fatigue strength exponent of −0.115. The fatigue life of the ASHM material is found to be dependent on the resulting porosity of the material that stems from the LW-DED process used in the ASHM process described. Full article

Accurately predicting the capacity degradation of lithium-ion batteries is crucial for ensuring the reliability and safety of electric vehicles and energy storage systems. However, existing methods—including those based on physical principles, deep learning, and traditional machine learning—all face challenges in balancing accuracy, computational efficiency, and adaptability to non-linear aging dynamics. This study proposes a new framework that combines multi-scale data preprocessing and a divide-and-conquer strategy to address these limitations. Firstly, a hybrid Wavelet–SG filter is applied to suppress noise, and a set of specialized XGBoost micro models is trained, with each model predicting capacity for a specific cycle, enabling precise trajectory prediction at different aging stages. The evaluation on the Toyota-MIT-Stanford dataset (118 batteries under different operating protocols) shows that this method achieves an average MAPE of 1.16% and a maximum of no more than 2.5% on the unfamiliar protocol test set. In terms of accuracy, it achieves performance comparable to CNN, LSTM, and CNN-LSTM benchmarks. Importantly, its parallel architecture enables fast inference (400 milliseconds on CPU), making it suitable for edge deployment in battery management systems. The model also has interpretability consistent with physical laws and can autonomously capture stage-dependent degradation mechanisms. This work provides a reliable, efficient, and interpretable solution for real-world battery health monitoring. Full article

Heap bioleaching is widely used to extract copper from low-grade sulfide ores thanks to its operational simplicity, low cost, and environmental sustainability. However, current control strategies rely primarily on single-factor optimization and often overlook the synergistic interactions of multiple key parameters, such as ore particle size, pore structure, pH, temperature, microbial activity, and oxygen transfer efficiency. As a result, issues such as low recovery rates, extended leaching periods, and high operational costs persist. Moreover, the “gray-box” nature of heap systems impedes real-time monitoring of internal physical, chemical, and biological processes. In addition, empirical multi-parameter optimization is time-consuming and inadequate for capturing complex interdependencies. This review was conducted to systematically examine the key factors influencing heap bioleaching efficiency and critically evaluate recent advances in numerical simulation and intelligent control strategies. As a result, we identified a major research gap: the existing models—including microscale shrinking core models (SCMs), mesoscale pore-network models based on CT reconstruction, and macroscale continuum models—have inherent limitations. SCMs assume idealized spherical particles with uniform mineral distribution while neglecting pore structure evolution and biofilm dynamics. Mesoscale models offer detailed pore characterization but lack robust multi-physics coupling (thermal–hydro–mechanical–chemical–biological, or THMCB). Macroscale models rely on homogenization assumptions that oversimplify spatial heterogeneity and temporal variations in permeability. This analysis covers the relevant literature from 1985 to 2025, with a focus on three methodological scales (micro, meso, and macro) and their integration with machine learning approaches. A notable finding is that hybrid neural network models (e.g., BP and RBF architectures) outperform purely physics-based models in predicting leaching kinetics under varying operational conditions. However, their accuracy depends heavily on high-quality field data—a limitation rarely addressed in prior reviews. By clearly delineating these model-specific limitations and scale-dependent trade-offs, this review makes two unique contributions: a structured framework for selecting and coupling numerical methods according to process requirements and a roadmap for integrating artificial neural networks with multi-physics simulations to achieve real-time intelligent control of heap bioleaching. The findings offer both theoretical guidance and practical references for optimizing the processing of low-grade copper sulfide ores. Full article

Micro- and nanoplastic (MNP) pollution has emerged as a global environmental and health concern, driving the rapid development of sensor technologies for faster, more sensitive, and field-deployable detection. This review synthesizes recent advances in optical, electrochemical, and biosensor platforms for MNP analysis and compares their analytical performance and practical feasibility. Optical sensors, including plasmonic, spectroscopic, and colorimetric systems, enable label-free and often rapid detection with material discrimination capability, and are well-suited for screening applications, though they commonly exhibit higher detection limits and matrix interference. Electrochemical sensors demonstrate the highest analytical sensitivity overall, frequently reaching low µg L⁻¹ to ng mL⁻¹ levels, with strong potential for miniaturization and on-site deployment; performance is further enhanced by nanostructured electrodes, photoelectrochemical designs, and signal amplification strategies. Biosensors incorporating peptides, aptamers, enzymes, or engineered proteins provide improved polymer selectivity and enable targeted detection, but face challenges related to stability, cross-reactivity, and reproducibility in complex samples. Practically, portable electrochemical and simple optical colorimetric platforms are currently the most feasible for field use, while hybrid bio-electrochemical systems show the highest performance potential. Future research should prioritize robust selective recognition elements, antifouling interfaces, standardized validation protocols, mixed-polymer quantification models, and integration with machine learning to enable reliable, real-world MNP monitoring. Full article

To address the issues of high cutting temperatures and severe tool wear during titanium alloy machining, this study proposes a hybrid surface modification strategy combining micro-textures and multicomponent titanium nitride (TiN)-based coatings on cemented carbide tools. Using YG8 cemented carbide as the substrate, micro-dimple textures were fabricated by fiber laser, and three coatings with different architectures (TiAlSiN, TiSiN/TiAlN, and TiSiN/TiAlSiN/TiAlN) were deposited via multi-arc ion plating technology. Based on a two-factor (texture diameter and texture spacing) and three-level orthogonal experiment, the evolution behaviors of surface morphology, phase composition, and mechanical properties of the textured multicomponent TiN-based coatings were systematically characterized and comparatively analyzed. The results reveal that: compared to the monolithic-structured TiAlSiN coating, the TiSiN/TiAlSiN/TiAlN and TiSiN/TiAlN composite coatings with multilayered composite structures can effectively relieve the residual stress inside the film–substrate system, and significantly suppress the phenomena of coating cracking and localized spallation caused by irregular protrusions of the recast layer at the micro-texture edges. X-ray diffraction (XRD) and crystallite size analyses indicate that the amorphous Si₃N₄ phase promoted by the Si element in the composite coatings effectively impedes the growth of TiN columnar crystals, achieving significant grain refinement. Mechanical property tests confirm that the existence of multicomponent composite interfaces effectively hinders dislocation movement. Among them, the textured TiSiN/TiAlSiN/TiAlN composite coating exhibits the optimal comprehensive performance; its microhardness, nanohardness, and H/E ratio (characterizing the resistance to plastic deformation) are increased by 17.94%, 8%, and approximately 45%, respectively, compared to those of the textured TiAlSiN coating. This study deeply elucidates the synergistic strengthening and toughening mechanisms between micro-texture parameters and the internal structures of the coatings, providing important theoretical guidance and experimental data support for the surface design of long-lifespan tools oriented towards the high-efficiency machining of titanium alloys. Full article

Machining Nomex honeycomb structures presents a major challenge due to their thin-walled architecture, orthotropic behavior, and sensitivity to adhesion and delamination. This study develops a three-dimensional numerical model using Abaqus/Explicit to analyze ultrasonic vibration-assisted milling in longitudinal and longitudinal-torsional modes. The model incorporates orthotropic behavior with progressive damage based on Tsai-Wu and experimental friction calibration to accurately reproduce tribological conditions. A parametric analysis examines the effect of vibration mode, amplitude (5–25 µm), frequency (21–22.5 kHz), cutting width, and tool geometry on stresses, bond wear, and material buildup. An optimal coefficient of friction ensures excellent simulation–experiment agreement. Compared to conventional milling, the longitudinal-torsional configuration reduces cutting forces by up to 50%, while frequency optimization allows for gains of 40 to 60%. Hybrid vibration coupling establishes intermittent contact and oscillatory micro-shearing, limiting adhesion and build-up. Thus, longitudinal-torsional assistance improves tribological stability, tool life and wall integrity, offering a validated digital strategy to optimize ultrasonic milling of composite honeycomb structures. Full article

Tubular flux-switching permanent-magnet linear machines (TFSPMLMs) are difficult to optimize using a single core material because conventional axial laminations suffer from severe in-plane eddy-current loss, whereas soft magnetic composites (SMCs) exhibit lower permeability and higher hysteresis loss. To address this trade-off, three hybrid SMC–laminated steel core configurations were investigated: H1, with radially laminated steel in the yoke; H2, with axially laminated steel in the tooth; and H3, with circumferential laminated steel segments. A reference SMC model (R1) and the three hybrid models were comparatively evaluated using three-dimensional finite element analysis (3D FEA). H1 and H2 showed degraded performance due to an interfacial micro-gap along the main flux path and additional in-plane eddy currents in the laminated steel regions. To mitigate these limitations, circumferential segmentation was applied to the laminated steel parts. With eight segments, H2 achieved a thrust force of 278.8 N, comparable to that of R1, while reducing iron loss by 22.5%; even a two-segment structure provided noticeable improvement. Among the investigated models, H3 showed the best overall performance by avoiding a micro-gap on the main flux path, achieving 285.5 N, and 3.9% higher thrust force and 18% lower iron loss than R1. These results indicate that H3 is the most effective hybrid-core configuration for maximizing both thrust force and loss reduction, whereas segmented H2 is an attractive practical option when manufacturability and low-loss operation are considered. Full article

Under dual-carbon targets, electric vehicles (EVs) have become central to transport decarbonization, making EV sales a key indicator of market diffusion and policy effectiveness. Despite the growing body of research, studies on EV sales remain fragmented and lack systematic integration. This study provides a structured review of EV sales research published between 2016 and 2025. Based on searches in Scopus and Web of Science, 1518 records were identified, and 194 peer-reviewed journal articles were retained after a multi-stage screening process. Temporal analysis reveals a clear stage-based evolution of EV sales research, with limited publications prior to 2020 and a marked expansion after 2021. The literature is categorized into two main streams: (i) determinants of EV sales and (ii) forecasting approaches. For determinants, a macro–meso–micro analytical framework is developed to organize policy, market, and behavioral factors. For forecasting, quantitative analysis shows that econometric and statistical models remain dominant (54%), while machine learning (18%), behavior simulation (14%), hybrid models (8%), and deep learning (4%) are increasingly adopted. This indicates a gradual shift toward data-driven and model integration approaches. This review offers a structured synthesis of determinant mechanisms and forecasting paradigms, identifies methodological imbalances, and outlines future research directions toward improved multi-level integration and mechanism-based modeling of EV sales dynamics. Full article

As network environments become increasingly interconnected, ensuring robust cyber-security has become critical, particularly with the growing sophistication of modern cyber threats. Intrusion detection systems (IDSs) play a vital role in identifying and mitigating unauthorized or malicious activities; however, conventional machine learning-based IDSs often rely on handcrafted features and are limited in their ability to detect diverse attack types across disparate network domains. To address these limitations, this paper introduces a novel unified intrusion detection framework that implements “Structural Dualism” to integrate three heterogeneous benchmark datasets (CSE-CIC-IDS2018, NF-BoT-IoT-v2, and IoT-23) into a harmonized, protocol-agnostic representation. The framework employs a shared autoencoder architecture with dataset-specific projection layers to learn a unified latent manifold. This 15-dimensional space captures the underlying semantics of attack patterns (e.g., volumetric vs. signaling) across multiple domains, while dataset-specific decoders preserve reconstruction fidelity through alternating multi-domain training. To identify complex micro-signatures within this manifold, the framework utilizes a synergistic hybrid convolutional neural network–deep neural network (CNN–DNN) classifier, where the CNN extracts spatial latent patterns and the DNN performs global classification across twenty-five distinct classes. Class imbalance is addressed through resampling strategies such as adaptive synthetic sampling (ADASYN) and edited nearest neighbors (ENN). Experimental results demonstrate remarkable performance, achieving 99.76% accuracy for binary classification and 99.54% accuracy for multi-class classification on the merged dataset, with strong generalization confirmed on individual datasets. These findings indicate that the shared autoencoder-based CNN–DNN framework, through its unique feature alignment and spatial extraction capabilities, significantly strengthens intrusion detection across diverse and heterogeneous environments. Full article

The widespread application of glyphosate—the world’s most used herbicide—presents a significant environmental challenge due to its persistence and mobility within interconnected land–soil–water systems. This study addresses the limitations of traditional, discrete water monitoring by developing a predictive framework for glyphosate and its primary metabolite, aminomethylphosphonic acid (AMPA), in the agricultural context of Apulia, Southern Italy. The methodology integrates high-sensitivity analytical chemistry with advanced spatial intelligence. Water samples were analyzed using an optimized UHPLC–MS/MS framework with pre-column derivatization (FMOC-Cl), achieving an ultra-trace Limit of Quantification (LOQ) of 0.025 μg/L. To transition from point data to continuous spatial profiles, a hybrid Machine Learning (ML) architecture was implemented. The model utilized a suite of geospatial predictors, including land use (Corine Land Cover), Digital Elevation Models (DEMs), and slope characteristics extracted from river offset lines. A dual-modeling strategy was employed: Global Models (Random Forest, Gradient Boosting, and KNN) for regional trends and Individual Models for river segments exhibiting sufficient internal variability. Analytical findings (2018–2024) revealed that AMPA consistently exhibited higher mean concentrations than glyphosate, reaching peaks of 9.27 μg/L. This trend is primarily attributed to its superior environmental persistence and a half-life of up to 240 days, compared to the parent compound. Spatiotemporal analysis identified critical peaks in the second quarter for glyphosate and extreme surges in the fourth quarter for AMPA, particularly in the Cervaro basin. The Random Forest Regressor emerged as the most robust predictive tool, achieving a coefficient of determination (R²) of approximately 0.68 at the global scale and up to 0.75 for localized models where data density was sufficient. The integration of ML frameworks allows for the identification of contamination “micro-hotspots” and the mapping of probabilistic pollutant distribution along entire river reaches without additional sampling costs. This high-fidelity diagnostic tool provides a cost-effective strategy for environmental agencies to implement targeted mitigation and proactive water resource protection in Mediterranean agroecosystems. Full article