Search Results (442)

The exponential increase in textual data generated across sectors such as healthcare, finance, and smart manufacturing has intensified the need for effective Big Data analytics. Large language models (LLMs) have become critical tools because of their advanced language processing capabilities. However, their static nature limits their ability to incorporate real-time and domain-specific knowledge. Retrieval-augmented generation (RAG) addresses these limitations by enriching LLM outputs through external content retrieval. Nevertheless, traditional RAG systems remain inefficient, often exhibiting high retrieval latency, redundancy, and diminished response quality when scaled to large datasets. This paper proposes an innovative structured RAG framework specifically designed for large-scale Big Data analytics. The framework transforms unstructured partial prompts into structured semantically coherent partial prompts, leveraging element-specific embedding models and dimensionality reduction techniques, such as principal component analysis. To further improve the retrieval accuracy and computational efficiency, we introduce a multi-level filtering approach integrating semantic constraints and redundancy elimination. In the experiments, the proposed method was compared with structured-format RAG. After generating prompts utilizing two methods, silhouette scores were computed to assess the quality of embedding clusters. The proposed method outperformed the baseline by improving the clustering quality by 32.3%. These results demonstrate the effectiveness of the framework in enhancing LLMs for accurate, diverse, and efficient decision-making in complex Big Data environments. Full article

Although evidence suggests adiposity as a modifiable risk factor for postmenopausal breast cancer (BC), its association with premenopausal BC remains uncertain. This potential differential relationship for menopausal status has been insufficiently investigated in the Moroccan population due to limited data. This study aims to assess the relationship between various indicators of adiposity and the risk of BC among Moroccan women by menopausal status. A multicenter case-control study was conducted in Morocco between December 2019 and August 2023, including 1400 incident BC cases and 1400 matched controls. Detailed measures of adiposity and self-reported measures from different life stages were collected. Unconditional logistic regression analyses were conducted to estimate odds ratios (ORs) and 95% confidence intervals (95% CIs) for the association between body size indicators and the risk of BC, adjusting for a range of known risk factors for BC. Higher waist circumference (WC) and hip circumference (HC) were associated with an increased risk of BC in both pre- (p-trend < 0.001 for both WC and HC) and post-menopausal women (p-trend < 0.001 for WC, 0.002 for HC). Current body mass index (BMI) ≥30 kg/m² increased the risk of postmenopausal BC (p-trend = 0.012). Among postmenopausal women, higher weight at age 20 was positively associated with BC risk (p-trend < 0.001), while, weight at age 30 was significantly associated with increased BC risk in both pre- (p-trend = 0.008) and post-menopausal women (p-trend = 0.028). Interestingly, weight gain since age 20 was inversely associated with BC risk in postmenopausal women in the adjusted model (p-trend = 0.006). Young-adult BMI observed a significant increased trend with BC risk in both pre- (p-trend = 0.008) and post-menopausal women (p-trend < 0.001). In premenopausal women, larger body shape during childhood and early adulthood was positively associated with BC risk (p-trend = 0.01 and = 0.011, respectively). In postmenopausal women, larger childhood and adolescent body silhouettes were also associated with increased BC risk (p-trend = 0.045 and 0.047, respectively). These results suggest that anthropometric factors may have different associations with pre- and post-menopausal BC among Moroccan women. This underscores the importance of conducting large prospective studies to better understand these findings and explore their links to different molecular subtypes of BC. Full article

While prior work has explored learner behavior using learning management systems (LMS) data, few studies provide week-level clustering validated against external engagement labels. To understand and assist students in online learning platforms and environments, this study presents a week-level engagement profiling framework for e-learning environments, utilizing K-means clustering and label-based validation. Leveraging log data from 127 students over a 13-week course, 44 activity-based features were engineered to classify student engagement into high, moderate, and low levels. The optimal number of clusters (k = 3) was identified using the elbow method and assessed through internal metrics, including a silhouette score of 0.493 and R² of 0.80. External validation confirmed strong alignment with pre-labeled engagement levels based on activity frequency and weighting. The clustering approach successfully revealed distinct behavioral patterns across engagement tiers, enabling a nuanced understanding of student interaction dynamics over time. Regression analysis further demonstrated a significant association between engagement levels and academic performance, underscoring the model’s potential as an early warning system for identifying at-risk learners. These findings suggest that clustering based on LMS behavior offers a scalable, data-driven strategy for improving learner support, personalizing instruction, and enhancing retention and academic outcomes in digital education settings such as MOOCs. Full article

This study evaluates hierarchical clustering methodologies to identify patterns associated with timeout requests for EuroLeague basketball games. Using play-by-play data from 3743 games spanning the 2008–2023 seasons (over 1.9 million instances), we applied Principal Component Analysis to reduce dimensionality and tested multiple agglomerative and divisive clustering techniques (e.g., Ward and DIANA) with different distance metrics (Euclidean, Manhattan, and Minkowski). Clustering quality was assessed using internal validation indices such as Silhouette, Dunn, Calinski–Harabasz, Davies–Bouldin, and Gap statistics. The results show that Ward.D and Ward.D2 methods using Euclidean distance generate well-balanced and clearly defined clusters. Two clusters offer the best overall quality, while four clusters allow for meaningful segmentation of game situations. The analysis revealed that teams that did not request timeouts often exhibited better scoring efficiency, particularly in the advanced game phases. These findings offer data-driven insights into timeout dynamics and contribute to strategic decision-making in professional basketball. Full article

This study proposes an application framework based on Large Language Models (LLMs) to analyze multimodal heterogeneous data in the power sector and introduces the CLB-BER model for classifying user electricity consumption behavior. We first employ the Euclidean–Cosine Dynamic Windowing (ECDW) method to optimize the adjustment phase of the CLUBS clustering algorithm, improving the classification accuracy of electricity consumption patterns and establishing a mapping between unlabeled behavioral features and user types. To overcome the limitations of traditional clustering algorithms in recognizing emerging consumption patterns, we fine-tune a pre-trained DistilBERT model and integrate it with a Softmax layer to enhance classification performance. The experimental results on real-world power grid data demonstrate that the CLB-BER model significantly outperforms conventional algorithms in terms of classification efficiency and accuracy, achieving 94.21% accuracy and an F1 score of 94.34%, compared to 92.13% accuracy for Transformer and lower accuracy for baselines like KNN (81.45%) and SVM (86.73%); additionally, the Improved-C clustering achieves a silhouette index of 0.63, surpassing CLUBS (0.62) and K-means (0.55), underscoring its potential for power grid analysis and user behavior understanding. Our framework inherently preserves temporal symmetry in consumption patterns through dynamic sequence alignment, enhancing its robustness for real-world applications. Full article

Gait recognition is a reliable biometric approach that uniquely identifies individuals based on their natural walking patterns. It is widely used to recognize individuals who are challenging to camouflage and do not require a person’s cooperation. The general face-based person recognition system often fails to determine the offender’s identity when they conceal their face by wearing helmets and masks to evade identification. In such cases, gait-based recognition is ideal for identifying offenders, and most existing work leverages a deep learning (DL) model. However, a single model often fails to capture a comprehensive selection of refined patterns in input data when external factors are present, such as variation in viewing angle, clothing, and carrying conditions. In response to this, this paper introduces a fusion-based multi-model gait recognition framework that leverages the potential of convolutional neural networks (CNNs) and a vision transformer (ViT) in an ensemble manner to enhance gait recognition performance. Here, CNNs capture spatiotemporal features, and ViT features multiple attention layers that focus on a particular region of the gait image. The first step in this framework is to obtain the Gait Energy Image (GEI) by averaging a height-normalized gait silhouette sequence over a gait cycle, which can handle the left–right gait symmetry of the gait. After that, the GEI image is fed through multiple pre-trained models and fine-tuned precisely to extract the depth spatiotemporal feature. Later, three separate fusion strategies are conducted, and the first one is decision-level fusion (DLF), which takes each model’s decision and employs majority voting for the final decision. The second is feature-level fusion (FLF), which combines the features from individual models through pointwise addition before performing gait recognition. Finally, a hybrid fusion combines DLF and FLF for gait recognition. The performance of the multi-model fusion-based framework was evaluated on three publicly available gait databases: CASIA-B, OU-ISIR D, and the OU-ISIR Large Population dataset. The experimental results demonstrate that the fusion-enhanced framework achieves superior performance. Full article

The precise identification of hydrocarbon-rich zones is crucial for optimizing exploration and production processes in the oil industry. Magnetotelluric (MT) surveys play a fundamental role in mapping subsurface geological structures. This study presents a novel methodology for automatically delineating resistivity contrasts in MT models by employing advanced machine learning and computer vision techniques. This approach commences with data augmentation to enhance the diversity and volume of resistivity data. Subsequently, a bilateral filter was applied to reduce noise while preserving edge details within the resistivity images. To further improve image contrast and highlight significant resistivity variations, contrast-limited adaptive histogram equalization (CLAHE) was employed. Finally, k-means clustering was utilized to segment the resistivity data into distinct groups based on resistivity values, enabling the identification of color features in different centroids. This facilitated the detection of regions with significant resistivity contrasts in the reservoir. From the clustered images, color masks were generated to visually differentiate the groups and calculate the area and proportion of each group within the pictures. Key features extracted from resistivity profiles were used to train unsupervised learning models capable of generalizing across different geological settings. The proposed methodology improves the accuracy of detecting zones with oil potential and offers scalable applicability to different datasets with minimal retraining, applicable to different subsurface environments. Ultimately, this study seeks to improve the efficiency of petroleum exploration by providing a high-precision automated framework with segmentation and contrast delineation for resistivity analysis, integrating advanced image processing and machine learning techniques. During initial analyses using only k-means, the resulting optimal value of the silhouette coefficient K was 2. After using bilateral filtering together with contrast-limited adaptive histogram equalization (CLAHE) and validation by an expert, the results were more representative, and six clusters were identified. Ultimately, this study seeks to improve the efficiency of petroleum exploration by providing a high-precision automated framework with segmentation and contrast delineation for resistivity analysis, integrating advanced image processing and machine learning techniques. Full article

The growing global demand for energy has resulted in a demand for innovative strategies for residential energy management. This study explores a novel framework—MELISSA (Modern Energy LLM-IoE Smart Solution for Automation)—that integrates Internet of Things (IoT) sensor networks with Large Language Models (LLMs) to optimize household energy consumption through intelligent automation and personalized interactions. The system combines real-time monitoring, machine learning algorithms for behavioral analysis, and natural language processing to deliver personalized, actionable recommendations through a conversational interface. A 12-month randomized controlled trial was conducted with 100 households, which were stratified across four socioeconomic quintiles in metropolitan areas. The experimental design included the continuous collection of IoT data. Baseline energy consumption was measured and compared with post-intervention usage to assess system impact. Statistical analyses included k-means clustering, multiple linear regression, and paired t-tests. The system achieved its intended goal, with a statistically significant reduction of 5.66% in energy consumption (95% CI: 5.21–6.11%, $p<0.001$) relative to baseline, alongside high user satisfaction (mean = 7.81, SD = 1.24). Clustering analysis ($k=4$, silhouette = 0.68) revealed four distinct energy-consumption profiles. Multiple regression analysis ($R^2=0.68$, $p<0.001$) identified household size, ambient temperature, and frequency of user engagement as the principal determinants of consumption. This research advances the theoretical understanding of human–AI interaction in energy management and provides robust empirical evidence of the effectiveness of LLM-mediated behavioral interventions. The findings underscore the potential of conversational AI applications in smart homes and have practical implications for optimization of residential energy use. Full article

The increasing complexity of financial reporting has enabled the implementation of innovative accounting practices that often obscure a company’s actual performance. This project seeks to uncover manipulative behaviours by constructing an anomaly detection model that utilises unsupervised machine learning techniques. We examined a dataset of 149,566 Slovak firms from 2016 to 2023, which included 12 financial parameters. Utilising TwoSteps and K-means clustering in IBM SPSS, we discerned patterns of normative financial activity and computed an abnormality index for each firm. Entities with the most significant deviation from cluster centroids were identified as suspicious. The model attained a silhouette score of 1.0, signifying outstanding clustering quality. We discovered a total of 231 anomalous firms, predominantly concentrated in sectors C (32.47%), G (13.42%), and L (7.36%). Our research indicates that anomaly-based models can markedly enhance the precision of fraud detection, especially in scenarios with scarce labelled data. The model integrates intricate data processing and delivers an exhaustive study of the regional and sectoral distribution of anomalies, thereby increasing its relevance in practical applications. Full article

Accurately reconstructing tree skeletons from multi-view images is challenging. While most existing works use skeletonization from 3D point clouds, thin branches with low-texture contrast often involve multi-view stereo (MVS) to produce noisy and fragmented point clouds, which break branch connectivity. Leveraging the recent development in accurate mask extraction from images, we introduce a mask-guided graph optimization framework that estimates a 3D skeleton directly from multi-view segmentation masks, bypassing the reliance on point cloud quality. In our method, a skeleton is modeled as a graph whose nodes store positions and radii while its adjacency matrix encodes branch connectivity. We use 3D Gaussian splatting (3DGS) to render silhouettes of the graph and directly optimize the nodes and the adjacency matrix to fit given multi-view silhouettes in a differentiable manner. Furthermore, we use a minimum spanning tree (MST) algorithm during the optimization loop to regularize the graph to a tree structure. Experiments on synthetic and real-world plants show consistent improvements in completeness and structural accuracy over existing point-cloud-based and heuristic baseline methods. Full article

Background: Lokomat-assisted robotic rehabilitation is increasingly used for gait restoration in patients with spinal cord injury (SCI). However, the objective evaluation of treatment effectiveness through biomechanical parameters and machine learning approaches remains underexplored. Methods: This study analyzed data from 29 SCI patients undergoing Lokomat-based rehabilitation. A dataset of 46 variables including range of motion (L-ROM), joint stiffness (L-STIFF), and muscular force (L-FORCE) was examined using statistical methods (paired t-test, ANOVA, and ordinary least squares regression), clustering techniques (k-means), dimensionality reduction (t-SNE), and anomaly detection (Isolation Forest). Predictive modeling was applied to assess the influence of age, speed, body weight, body weight support, and exercise duration on biomechanical outcomes. Results: No statistically significant asymmetries were found between left and right limb measurements, indicating functional symmetry post-treatment (p > 0.05). Clustering analysis revealed a weak structure among patient groups (Silhouette score ≈ 0.31). Isolation Forest identified minimal anomalies in stiffness data, supporting treatment consistency. Regression models showed that body weight and body weight support significantly influenced joint stiffness (p < 0.01), explaining up to 60% of the variance in outcomes. Conclusions: Lokomat-assisted robotic rehabilitation demonstrates high functional symmetry and biomechanical consistency in SCI patients. Machine learning methods provided meaningful insight into the structure and predictability of outcomes, highlighting the clinical value of weight and support parameters in tailoring recovery protocols. Full article

Background/Objectives: A faster eating rate is associated with increased energy intake and risk of obesity. High consumption of ultra-processed foods (UPFs) is associated with a faster eating rate. Psychobehavioral aspects, such as body image self-perception, eating disorders, and anxiety, may modulate this eating behavior. Therefore, this study examined the moderating role of psychobehavioral factors in the association between meal type (UPF vs non-UPF) and eating rate among individuals with obesity. Methods: It is a secondary analysis of a randomized, parallel clinical trial conducted with 39 adults who have obesity. Participants were assigned to consume either a UPF-only composed meal or a UPF-free meal, both of which were isoenergetic (~550 kcal). Psychobehavioral variables (food addiction—mYFAS 2.0, body image perception and satisfaction—Silhouette Rating Scale, eating disorders—EAT-26, and anxiety—GAD-7) were assessed. Eating rate was measured in kcal/min. Results: Body image perception and satisfaction significantly interacted with the type of meal. In the UPF group, lower body image dissatisfaction was associated with a higher eating rate (β = 4.79 kcal/min; 95% CI: 1.40; 8.19; p = 0.007), while a higher body image perception score was associated with a lower eating rate (β = −4.61 kcal/min; 95% CI: −8.57; −0.65; p = 0.024). No significant associations were observed for food addiction scores, eating disorders or anxiety. Conclusions: Body image modulates the eating rate in the context of UPF consumption. These findings suggest that interventions against obesity should consider individual psychobehavioral characteristics. Full article

Effective monitoring of fish passage through river barriers is essential for evaluating fishway performance and supporting adaptive river management. Traditional methods are often invasive, labor-intensive, or too costly to enable widespread implementation across most fishways. Infrared (IR) beam-break counters offer a promising alternative, but their adoption has been limited by high costs and a lack of flexibility. We developed and tested a novel, low-cost infrared beam-break counter—FishTracker—based on open-source Raspberry Pi and Arduino platforms. The system detects fish passages by analyzing interruptions in an IR curtain and reconstructing fish silhouettes to estimate movement, direction, speed, and morphometrics under a wide range of turbidity conditions. It also offers remote access capabilities for easy management. Field validation involved controlled tests with dummy fish, experiments with small-bodied live specimens (bleak) under varying turbidity conditions, and verification against synchronized video of free-swimming fish (koi carp). This first version of FishTracker achieved detection rates of 95–100% under controlled conditions and approximately 70% in semi-natural conditions, comparable to commercial counters. Most errors were due to surface distortion caused by partial submersion during the experimental setup, which could be avoided by fully submerging the device. Body length estimation based on passage speed and beam-interruption duration proved consistent, aligning with published allometric models for carps. FishTracker offers a promising and affordable solution for non-invasive fish monitoring in multispecies contexts. Its design, based primarily on open technology, allows for flexible adaptation and broad deployment, particularly in locations where commercial technologies are economically unfeasible. Full article

This paper presents a novel approach to supplier clustering by utilizing the Arithmetic Optimizer Algorithm (AOA), addressing the complex challenge of supplier segmentation in modern supply chain management. The AOA framework is applied to solve the multi-criteria clustering problem inherent to supplier classification. Using a real-world dataset of 500 suppliers with 12 performance criteria, including cost, quality, delivery reliability, and sustainability metrics, our method demonstrates effective clustering performance compared to conventional techniques. The AOA achieves a silhouette coefficient of 56.5% and a Davies–Bouldin index of 56.6%, outperforming several other state-of-the-art metaheuristic algorithms, including the Grey Wolf Optimizer (GWO), Whale Optimization Algorithm (WOA), Salp Swarm Algorithm (SSA), and Harris Hawks Optimization (HHO). The algorithm’s robustness is validated through extensive sensitivity analysis and statistical tests. The results indicate that the proposed approach successfully identifies distinct supplier segments with approximately 85% accuracy, enabling more effective supplier relationship management strategies. Full article

The safe and stable operation of hydropower units, as core equipment in clean energy systems, is crucial for power system security. However, anomaly detection under complex operating conditions remains a technical challenge in this field. This paper proposes a hydropower unit anomaly detection method based on K-means seeded quadratic discriminant clustering and an adaptive density-aware ensemble anomaly detection algorithm (KSQDC-ADEAD). The method first employs the KSQDC algorithm to identify different operating conditions of hydropower units. By combining K-means clustering’s initial partitioning capability with quadratic discriminant analysis’s nonlinear decision boundary construction ability, it achieves the high-precision identification of complex nonlinear condition boundaries. Then, an ADEAD algorithm is designed, which incorporates local density information and improves anomaly detection accuracy and stability through multi-model ensemble and density-adaptive strategies. Validation experiments using 14-month actual operational data from a 550 MW unit at a hydropower station in Southwest China show that the KSQDC algorithm achieves a silhouette coefficient of 0.64 in condition recognition, significantly outperforming traditional methods, and the KSQDC-ADEAD algorithm achieves comprehensive scores of 0.30, 0.34, and 0.23 for anomaly detection at three key monitoring points, effectively improving the accuracy and reliability of anomaly detection. This method provides a systematic technical solution for hydropower unit condition monitoring and predictive maintenance. Full article