Search Results (812)

This study aims to develop an efficient and accurate method for identifying joint parameters in assembled structures. A novel neural network-based joint identification framework is proposed. Frequency response function (FRF) datasets are generated by combining finite element simulation with frequency-domain substructure synthesis. The Uniform Manifold Approximation and Projection (UMAP) algorithm is employed for nonlinear dimensionality reduction in FRF sequences, preserving critical characteristics. A multilayer perceptron (MLP) network is then trained to regress joint parameters from the reduced-dimension FRF data. The necessity of the nonlinear dimensionality reduction within this joint identification framework is verified through comparison with the linear dimensionality reduction technique of principal component analysis (PCA). This methodology is implemented and validated using a tool–collet–holder system. Comparative studies with the global optimization method reveal that the proposed approach maintains superior identification accuracy while achieving significant improvements in computational efficiency across varying preload conditions. Furthermore, the identified joint parameters exhibit strong predictive capability when tested under tool/holder component changes, preload variations, and when coupled with a spindle, proving robustness under complex operational scenarios. This study provides a new technical pathway for the joint identification of assembly structure. Full article

Oxalic acid is extensively used in industrial chemical processes, purification systems, hydrometallurgical operations, and advanced oxidation environments where rapid and environmentally sustainable analytical methodologies are increasingly required for process monitoring and quality control. In this study, a micro-Raman spectroscopy methodology was developed for the direct quantification of oxalic acid in aqueous systems at moderate-to-high concentrations (0.079–0.793 M). The analytical strategy was based on the integrated Raman response of the carbonyl stretching region (1700–1750 cm⁻¹), selected due to its strong concentration-dependent behavior, spectral definition, and reduced interference from the aqueous matrix. The proposed methodology demonstrated excellent analytical performance, including high linearity (R² > 0.998), satisfactory precision, and reliable concentration-dependent reproducibility throughout the evaluated concentration range. To evaluate operational robustness, matrix-matched standards incorporating temperature variation (25–40 °C), turbidity (0–57 mg/L), dissolved Ca²⁺ (0–58 mg/L), and dissolved Fe³⁺ (0–7 mg/L) were prepared to simulate chemically perturbed industrial environments. Principal Component Analysis (PCA) demonstrated that the carbonyl vibrational region retained organized concentration-dependent spectral behavior despite operational perturbations. Partial Least Squares (PLS) regression models developed under these matrix-informed conditions preserved strong predictive capability (R² ≈ 0.997), while preliminary prediction of process-related samples yielded excellent agreement between predicted and reference concentrations (R² = 0.990). Although operational perturbations produced substantial attenuation of Raman intensity, particularly at lower concentration levels, the carbonyl Raman band remained spectrally detectable and analytically interpretable throughout all evaluated conditions. Electronic-structure analysis using Natural Bond Orbital (NBO) and Atoms-in-Molecules (AIM) methodologies demonstrated that the strong analytical behavior of the ν(C=O) vibrational mode is associated with enhanced electron-density localization, covalent stabilization, and favorable polarizability characteristics of the carbonyl bond. The combined experimental, chemometric, and computational results demonstrate the feasibility of matrix-informed micro-Raman spectroscopy as a rapid, reagent-free, and operationally robust methodology for oxalic acid monitoring in chemically perturbed aqueous industrial systems. Full article

Quantum machine learning on near-term noisy quantum devices has generated substantial theoretical interest, but rigorous empirical comparisons under realistic noise on practically relevant data remain scarce. This paper compares two paradigmatic quantum learning models, a Quantum Support Vector Machine (QSVM) built on the ZZFeatureMap quantum kernel and a Variational Quantum Classifier (VQC) with an EfficientSU2/RealAmplitudes ansatz, against tuned classical baselines (SVM with four kernels, Random Forest, XGBoost, LightGBM and CatBoost) on the ULB Credit Card Fraud dataset (284,807 transactions, 0.17% fraud). All models share an identical 4-qubit PCA-reduced feature space, evaluated on the full unbalanced test fold over 15 fits (3 folds × 5 seeds) and reported as mean ± standard deviation with bootstrap confidence intervals, AUPRC as the primary metric. Noise robustness is assessed under depolarizing noise $p\in \{0,0.001,0.01,0.05\}$, with ranking preservation measured directly through Spearman $\rho$ and Kendall $\tau$ between the noisy and noiseless decision scores rather than read off AUPRC, alongside the per-paradigm computational cost. At four qubits the classical baselines lead (AUPRC 0.60 to 0.74, CatBoost best), above the VQC (0.494) and the QSVM (0.240); the controlled QSVM-versus-RBF–SVM comparison puts the cost of the quantum kernel at about 0.45 AUPRC. Under noise the QSVM keeps its score ranking ($\rho =0.998$ at $p=0.001$, 0.906 at $p=0.01$) and an operational decision threshold (recall 0.87 to 0.89, stable calibration), while the VQC AUPRC peaks non-monotonically at $p=0.01$ (0.494 rising to 0.654, then easing to 0.569 at $p=0.05$) even as its ranking decays monotonically ($\rho$ from 0.72 to near zero), so average precision on its own misrepresents how noise affects it. The quantum models do not surpass the tuned classical reference at four qubits; the contribution is methodological: under noise, AUPRC has to be read together with a genuine rank statistic, because the two can move in opposite directions. Full article

Background: Ischemic cardiomyopathy (ICM) is characterized by heterogeneous structural remodeling that is not fully captured by conventional systolic metrics. How multidimensional structural echocardiographic information can improve pre-revascularization risk stratification remains unclear. Methods: In this retrospective study, 989 patients with ICM undergoing coronary angiography and revascularization were included in the derivation cohort, and 482 patients from an independent campus served as the validation cohort, with a median follow-up duration of 6.5 years. The primary endpoint was cardiovascular mortality. Eight routinely acquired pre-revascularization echocardiographic structural variables were analyzed. Unsupervised clustering identified structural clusters, and principal component analysis (PCA) was used to derive a structural risk score. Associations with cardiovascular mortality were assessed using the Cox proportional hazards model, and prognostic performance was evaluated by comparing individual echocardiographic predictors using Harrell’s C-index and time-dependent AUC analyses. Results: Three distinct structural clusters emerged, differing in chamber size, systolic function, pulmonary pressures, mitral regurgitation severity, and long-term cardiovascular mortality. The PCA-derived structural risk score, reflecting the dominant axis of remodeling and volume overload, showed association with cardiovascular mortality in the derivation cohort and remained independently predictive after multivariable adjustment. Compared with single echocardiographic parameters, both the structural clusters and the risk score demonstrated superior discriminative performance. In the validation cohort, the structural score again showed a consistent and independent association with cardiovascular mortality. Conclusions: Multidimensional structural echocardiographic assessment reveals clinically meaningful remodeling patterns and enables construction of a robust PCA-derived structural risk score. Both approaches provide prognostic information beyond individual echocardiographic measures and support more precise pre-revascularization risk stratification in patients with ICM. Full article

Extracting reliable damage-sensitive features from sparse sensor networks under Environmental and Operational Variations (EOV) remains a critical challenge in Structural Health Monitoring (SHM). The purpose of this study is to overcome this limitation by proposing a novel, data-driven framework utilizing a cost-effective network of high-sensitivity triaxial roving accelerometers. The methodology integrates an AutoRegressive with eXogenous inputs (ARX) model and Wavelet Packet Decomposition (WPD) to extract robust, damage-sensitive features from complex vibration data. To handle the high-dimensionality of the extracted signals and achieve optimal multi-sensor data fusion, Block-wise Principal Component Analysis (PCA) is employed as a signal sanitation and feature reduction tool. This algorithmic pipeline is applied to a full-scale bridge pier subjected to RC jacketing. The structural enhancements and dynamic behavior shifts post-retrofitting were statistically quantified using the Mahala Nobis distance. The analysis revealed a 41.2% attenuation in median vibration intensity and successfully verified the structural improvements at a 99% confidence interval, clearly distinguishing the retrofitting effects from ambient noise. The proposed framework successfully isolates true structural changes from EOV, providing a reliable non-destructive evaluation tool for continuous monitoring in practical civil engineering applications. Full article

Air pollution is primarily caused by human activities such as industrial emissions, road traffic, waste incineration, and fossil fuel power plants. Pollution refers to the presence of harmful substances in the air, such as nitrogen dioxide (NO₂), sulfur dioxide (SO₂), ozone (O₃), carbon monoxide (CO), and other environmental pollutants. Some pollutants pose health risks even at low doses. Given the critical importance of air quality, monitoring air pollution has become an urgent and essential subject. Air quality monitoring relies on accurate data, so changeable environments and sensor issues make using interval diagnostic techniques for addressing uncertainty in systems interesting. In this article, we focus on three key aspects to achieve precise and efficient results: (1) the use of an accurate fault detection method that accounts for data uncertainty while maintaining model symmetry, (2) the implementation of a reliable detection index invariant to symmetric sensor behaviors, and (3) the combination of both to improve fault localization accuracy. This paper presented a fault detection and localization framework designed for uncertain and nonlinear monitoring environments. A novel fault-sensitive detection index was developed and integrated into an elimination-based localization strategy within a reduced-rank interval kernel PCA (RR-IKPCA) model. By exploiting information contained in modified residual subspaces and explicitly accounting for measurement uncertainty, the proposed approach enhances fault sensitivity while preserving robust localization capability, as validated on the AIRLOR air quality monitoring network. Full article

The rapid development of the Industrial Internet of Things (IIoT) generates massive heterogeneous sensor data, complicating data cleaning and normalization. Existing algorithmcentric methods often treat quality issues in isolation and lack unified governance. This paper proposes a governance-centered framework for multi-source industrial sensor data. We introduce an Intelligent Catalog as the semantic governance layer to standardize metadata and achieve semantic alignment before numerical processing. Building upon this, an AI Agent-driven mechanism dynamically orchestrates cleaning and normalization strategies based on real-time data status and heterogeneous features. This framework modularly integrates classical algorithms (e.g., PCA, KPCA, LSTM) without model dependency. Experimental results on public IIoT datasets demonstrate that our framework significantly outperforms baseline methods in normalization consistency, noise robustness, and stability across heterogeneous data. By shifting from an algorithm-centered to a governance-centered paradigm, this approach provides a scalable and adaptive solution for complex industrial sensor data management. Full article

Individual tree classification has a long history of diverse development, with recent trends focusing on the adoption of machine learning and deep learning approaches. It is a simple and powerful approach that allows the model to auto-pilot while reducing the need for physical characteristic understanding. Over more than a decade of research, we have focused on establishing a direct representation of individual trees that bridges 2D top-down imagery and true 3D models. In this study, we investigated the fundamental question of the influence of the input data on these ML/DL models. In 2024, we introduced a novel data transformation method, the Pseudo Tree Crown (PTC), which provides a pseudo-3D pixel-value perspective that enhances the informational richness of images and significantly improves classification performance. Our original implementation was successfully tested on urban and deciduous trees in 2024 and was later extended to Canadian natural conifer species under snow conditions in 2025. However, the original PTC relied on the green band, limiting its applicability to green-leaf species. In this study, we analyzed and compared the performance of different data variations and transformations, such as the Green–Red Vegetation Index (GRVI) and principal component analysis (PCA), as direct input and used their PTC forms. Classifications were conducted using Random Forest (RF), ResNet50, YOLOv10 and Segment Anything (SA). The results confirmed the effectiveness of the PTC, which consistently improves the classification accuracy by at least 5% without introducing additional computational time or complexity. Furthermore, PTC exhibits robust, consistent behavior across all data forms, demonstrating its strong resilience and reliability. Full article

Landslide detection has largely relied on supervised learning or DEM-based representations, which can limit rapid deployment and generalization across heterogeneous terrain. In this work, we present a zero-shot, fully unsupervised framework that identifies landslide-like geomorphic instability candidates from raw UAV-mounted LiDAR, removing the need for labeled data, pre-event baselines, or rasterized terrain abstractions. Our approach is motivated by the observation that landslides manifest as localized geometric inconsistencies in the terrain surface. We capture this through a multi-scale formulation that combines point-level and cluster-level indicators of instability. At the point level, a PCA-based residual depth metric reduces slope-induced bias and highlights surface discontinuities, while local concavity captures terrain depletion patterns. At the cluster level, geomorphometric descriptors such as curvature concentration, surface roughness, elevation discontinuity, and slope variation are extracted using density-aware 3D clustering and integrated through adaptive feature fusion. The resulting probabilistic instability field enables spatially coherent delineation of landslide scars, including rupture boundaries, displaced material, and emerging failure regions. In addition, the detected patches provide useful priors for post-event susceptibility analysis without requiring temporal observations. Experiments across diverse geomorphic settings show that the proposed method improves detection of subtle terrain disturbances compared to DEM-based pipelines and supervised learning approaches, while remaining robust to noise and terrain variability. Overall, this work demonstrates that geometry-driven, unsupervised inference on raw 3D data can serve as a practical and scalable alternative for near real-time landslide detection using UAV-based systems. Full article

Interpreting complex, non-linear Interferometric Synthetic Aperture Radar (InSAR) displacement time-series data for infrastructure risk assessment remains a significant geotechnical challenge. This is particularly evident in regions with established road and railway infrastructures, where the primary objective is monitoring the entire network to ensure safety and operational continuity. Because landslide displacement is a highly complex process affected by a combination of internal geological conditions and external triggers, time-series data inherently encode non-linear trends and periodic fluctuations. To address this, a data-driven framework utilizing a sliding-window transformation to engineer temporal-kinematic features is proposed, providing a broader framework for the contextualization of slope stability assessment from a network perspective. This is paired with Principal Component Analysis (PCA) for dimensionality reduction and evaluated across four unsupervised architectures: K-means, Self-Organising Maps (SOMs), Hybrid SOM-K-means, and PCA-K-means. The comparative evaluation reveals that the PCA-K-means pipeline performed best, offering a highly efficient and scalable workflow. The analysis revealed that the optimized PCA-K-means architecture successfully captured 79.20% of the kinematic variance across the first two principal components. Furthermore, it achieved a robust Between-Cluster-to-Total-Sum-of-Squares (BCSS/TSS) ratio of 71.70%, an optimal Silhouette Score of 0.320, and a low Quantisation Error (QE) of 0.90, demonstrating superior spatial separation and geometric accuracy compared to traditional heuristic methods. When cross-validated against static topographic susceptibility models, the dynamic kinematic clusters exhibited a 23% spatial convergence at the polar bounds of risk, successfully grounding the algorithm’s predictions in physically verified geomorphological features. Relying on the statistical volatility of displacements, this optimal model successfully partitioned the data into five distinct geotechnical risk classes, ranging from stable (Class A) to extreme risk (Class E). The results demonstrate that the developed dynamic framework provides a highly reliable, actionable tool for proactive, large-scale slope stability and infrastructure risk assessment. Full article

Biflavonoids are dimeric flavonoids recognized for their diverse biological activities and significant pharmacological potential, with ginkgo (Ginkgo biloba L.) serving as a primary natural source. This study presents a comprehensive spatiotemporal characterization of the biflavonoid profile across a diverse population of 90 trees. High-resolution chromatographic analysis quantified five major biflavonoids, revealing a consistent hierarchical abundance: sciadopitysin > isoginkgetin > ginkgetin > bilobetin > amentoflavone. Notably, sciadopitysin emerged as the predominant constituent (1532.89 ± 544.13 µg/g dw). To decode the complex drivers of metabolite accumulation, we integrated Principal Component Analysis (PCA) with Piecewise Linear Regression (PLR). PCA confirmed a robust chemical structure, explaining 71.5% of the total variance, where Factor 1 represents a general biflavonoid gradient and Factor 2 captures localized environmental influences. The PLR models (R² = 0.75–0.83) identified tree age as a primary negative regulator, showing a significant decline in total biflavonoids as trees mature beyond the 30-year reproductive threshold. While sexual dimorphism and location exhibited compound-specific nonlinear effects, younger trees (10–30 years) demonstrated the highest biosynthetic plasticity and potency. These findings establish a predictive framework for optimizing the pharmaceutical harvest of ginkgo leaves, highlighting that age-related physiological shifts, rather than gender or broad geography, are the critical determinants of biflavonoids yield. Full article

To address the limited cross-domain generalization of single-algorithm models for shield attitude prediction, this study proposes a heterogeneous algorithm-fusion framework based on Stacking. The framework integrates multiple machine learning algorithms and uses the Spider Wasp Optimizer (SWO) for hyperparameter optimization, thereby overcoming the limitations of individual learners and reducing the need for laborious algorithm selection. Principal Component Analysis (PCA) is further used to reduce dimensionality and reconstruct high-dimensional features, which lowers computational complexity and improves prediction accuracy. The proposed PCA-SWO-Stacking algorithm was applied to shield attitude prediction using data from the Shanghai Beiheng Passageway project. The results show strong predictive performance, with all R² values exceeding 0.94 and all RMSE and MAE values remaining below 2. Comparative experiments with commonly used ensemble algorithms and ablation studies further confirm the effectiveness and robustness of the proposed method. Full article

Cognitive stress detection based on electrocardiogram (ECG) is challenged by the high dimensionality of multichannel analysis, redundancy between heart rate variability (HRV) and morphological descriptors, and variability in classifier performance. We developed and evaluated a cognitive stress classification framework based on a standardized ECG acquisition protocol, the integration of HRV and morphological descriptors extracted from 12 leads, and an adaptive feature selection strategy using a binary genetic algorithm with explicit penalization of dimensionality. Seventy healthy students aged 18–25 years participated, and cognitive stress was induced using a task based on PMA-R Factor R. The initial dataset included 27 descriptors per lead, and the proposed dimensionality reduction method was compared with two reference schemes: no dimensionality reduction and conventional principal component analysis (PCA) with a 99% cumulative explained variance threshold. Performance was assessed over 30 data splits using five classifiers: logistic regression, linear support vector machine (SVM), radial basis function SVM (SVM-RBF), k-nearest neighbors (KNN), and decision tree. The best trade-off between parsimony and predictive performance was achieved with $\lambda =0.05$, yielding a compact subset of 11 features on average and a mean AUC of 0.830. In the final comparison, the adaptive strategy achieved the best overall performance with SVM-RBF (AUC $=0.830\pm 0.047$; specificity $=0.814\pm 0.115$), outperforming both the full feature set and PCA. These findings indicate that penalized genetic selection validated across multiple classifiers is an effective strategy for identifying compact, discriminative, and robust feature subsets for ECG-based cognitive stress classification. Full article

Prostate cancer (PCa) remains a major global health concern, with a subset of patients progressing to aggressive disease despite advances in diagnosis and treatment. Epithelial–mesenchymal transition (EMT) plays a pivotal role in tumor invasion, metastasis, and immune evasion; however, its cellular heterogeneity and clinical relevance in PCa remain incompletely understood. We analyzed single-cell transcriptomic data to characterize EMT dynamics in malignant epithelial cells. Malignant cells were identified based on aberrant copy number variation patterns, and EMT activity was quantified using AUCell. Gene expression profiling and gene set enrichment analysis identified key EMT-associated genes. By integrating bulk transcriptomic data with LASSO regression analysis, we identified five pivotal genes and constructed an EMT infiltration scoring model. The model demonstrated robust predictive performance in an external Gene Expression Omnibus validation cohort and effectively predicted early biochemical recurrence. Further analyses revealed significant associations between EMT scores, clinicopathological features, immune cell infiltration, genomic instability, and tumor immune dysfunction and exclusion scores. Pathway enrichment analysis highlighted distinct molecular characteristics between high- and low-score groups. Additionally, molecular docking using AutoDock identified potential targeted therapeutic agents for key EMT genes. Overall, this study systematically delineates EMT heterogeneity at the single-cell level and establishes a robust EMT infiltration model for prognostic prediction and therapeutic guidance in PCa, providing novel insights for precision risk stratification and individualized treatment strategies. Full article

Pumpkin (Cucurbita spp.) is a globally significant vegetable crop known for its high nutritional value and remarkable phenotypic diversity. Yet, the surge in new cultivar releases has overwhelmed traditional morphological descriptors, creating critical gaps in variety purity control and breeders’ rights enforcement. Despite the established utility of SNP markers as the gold standard for genetic analysis, a dedicated high-resolution molecular database for modern pumpkin cultivars remains unavailable. To address this gap, we conducted whole-genome resequencing (WGS) on 94 representative pumpkin cultivars (spanning C. moschata, C. maxima, and C. pepo). Clean reads were mapped to the Cucurbita maxima reference genome. We employed a stringent pipeline to identify genomic variants and utilized STRUCTURE software, Principal Component Analysis (PCA), and Neighbor-Joining (NJ) trees to evaluate population stratification. Linkage disequilibrium (LD) decay and DNA fingerprinting barcodes were also developed. A total of 8,873,150 high-quality variants were identified, including 7,345,007 SNPs and 1,528,143 InDels, with an average SNP density of 21,281.50 SNPs/Mb. Population analysis consistently categorized the 94 cultivars into two primary subpopulations (G1 and G2). The first two PCs accounted for 74.06% of the total genetic variance. Further analysis revealed that G1 possessed a more complex genetic architecture and slower LD decay compared to G2, suggesting distinct selection histories. Finally, we screened for highly informative biallelic SNPs to construct a DNA fingerprinting database, enabling precise sample discrimination through unique chromatic barcodes. This study fills a critical gap in pumpkin genomics by establishing a high-density SNP database and a robust fingerprinting system. These resources provide a definitive tool for variety certification, seed purity testing, and the advancement of molecular-assisted breeding in pumpkin. Full article