Search Results (6,590)

Recent advancements in wearable technology have allowed researchers to collect high-resolution data on athletes’ workloads and performance, paving the way for more accurate mathematical models in sports science. In this paper, inspired by the modeling of heart rate during exercise, we introduce a novel linear, time-varying, two-state discrete-time dynamical model for predicting athletic performance in elite-level soccer players. Model parameters are estimated via the Differential Evolution optimization algorithm, and GPS-derived metrics such as metabolic power and equivalent distance index are incorporated. The model originally accounts for complex interactions between a performance-related state variable and a second lumped variable, whose dynamics are intertwined. This model was compared to the most effective deterministic (though uncertain) one in the literature, namely the (nonlinear) Busso model. Results, concerning two professional soccer players over a half-season period, show that the proposed model outperforms the traditional approach in estimation and predictive accuracy, with significantly higher correlation coefficients and lower estimation and prediction errors across all players. These findings suggest that integrating two-state dynamics and fine-grained GPS metrics provides a more biologically realistic framework for load monitoring in team sports. The proposed model thus represents a powerful tool for training optimization and athlete readiness assessment, with potential applications in real-time decision support systems for coaching staff. By predicting the effects of training load on future performance, it might also contribute to injury risk reduction and the prevention of maladaptive responses to excessive workload. Full article

Auxiliary systems, particularly HVAC and thermal management, significantly influence electric vehicle (EV) range under diverse weather conditions. Accurate prediction of auxiliary power demand remains challenging due to nonlinear temperature dependencies and driving dynamics. Here we develop an integrated physics-based decomposition combined with an XGBoost machine learning model trained on 95,028 real-world measurements from EVs operating across multi-seasonal conditions (−8 °C to +33.5 °C). The model achieves an R² of 0.9986 and a mean absolute error of 35 W, revealing that auxiliary loads contribute variably from 75% while idle to 12% during highway driving, with heating power dominating cooling by a 7:1 ratio and increasing 44-fold at low temperatures. Feature importance analysis identifies accelerator pedal position and heating efficiency per temperature differential as primary predictors, indicating coupling between propulsion and auxiliary loads. These findings underscore the necessity of context-aware auxiliary power prediction to enhance EV energy management and range forecasting, particularly in cold climates where heating demands critically impact efficiency. Full article

Background: Tacrolimus is a cornerstone immunosuppressant in kidney transplantation, but its narrow therapeutic index and marked inter-patient variability complicate dose optimization. Conventional therapeutic drug monitoring (TDM) relies on empirical adjustments that often overlook individual pharmacokinetics. Machine learning (ML) offers a precision dosing alternative by integrating diverse clinical and biochemical variables into predictive models. Methods: We retrospectively analyzed 1351 data points from 87 kidney transplant patients at Eunpyeong St. Mary’s Hospital (April 2019–November 2023). Clinical, demographic, and laboratory information, including tacrolimus trough levels and dosing history, were extracted from electronic medical records. Four predictive models—XGBoost, CatBoost, LightGBM, and a multilayer perceptron (MLP)—were trained to forecast next-day tacrolimus concentrations, and model serum creatinine level performance was evaluated using R-squared (R²), mean absolute error (MAE), and root-mean-squared error (RMSE). An ensemble model with weighted soft voting was applied to enhance predictive accuracy, and model interpretability was assessed using SHapley Additive exPlanations (SHAP). Results: The ensemble model achieved the best overall performance (R² = 0.6297, MAE = 1.0181, RMSE = 1.2999), outperforming all individual models, whereas the MLP model showed superior predictive power among single models, reflecting the significance of nonlinear interactions in tacrolimus pharmacokinetics. SHAP analysis highlighted prior tacrolimus levels, cumulative dose, renal function markers (eGFR level, serum creatinine level), and albumin concentration as the most influential predictors. Conclusions: We present a robust ML-based algorithm for tacrolimus dose optimization in hospitalized kidney transplant recipients. By improving predictions of tacrolimus concentrations, the model may help reduce inter-patient dose variability and lower the risk of nephrotoxicity, supporting safer and more individualized immunosuppressive management. This approach advances AI-driven precision medicine in transplant care, offering a pathway to safer and more effective immunosuppression. Full article

Adversarial patch attacks have emerged as a powerful and practical threat to machine learning models in vision-based tasks. Unlike traditional perturbation-based adversarial attacks, which often require imperceptible changes to the entire input, patch attacks introduce localized and visible modifications that can consistently mislead deep neural networks across varying conditions. Their physical realizability makes them particularly concerning for real-world security-critical applications. In response, a growing body of research has proposed diverse defense strategies, including input preprocessing, robust model training, detection-based approaches, and certified defense mechanisms. In this paper, we provide a comprehensive review of patch-based adversarial attacks and corresponding defense techniques. First, we introduce a new task-oriented taxonomy that systematically categorizes patch attack methods according to their downstream vision applications (e.g., classification, detection, segmentation), and then we summarize defense mechanisms based on three major strategies: Patch Localization and Removal-based Defenses, Input Transformation and Reconstruction-based Defenses, Model Modification and Training-based Defenses. This unified framework provides an integrated perspective that bridges attack and defense research. Furthermore, we highlight open challenges, such as balancing robustness and model utility, addressing adaptive attackers, and ensuring physical-world resilience. Finally, we outline promising research directions to inspire future work toward building trustworthy and robust vision systems against patch-based adversarial threats. Full article

The Internet of Things (IoT) has rapidly evolved into a central component of today’s technological landscape, enabling seamless connectivity and communication among a vast array of devices. It underpins automation, real-time monitoring, and smart infrastructure, serving as a foundation for Industry 4.0 and paving the way toward Industry 5.0. Despite the potential of IoT systems to transform industries, these systems face a number of challenges, most notably the lack of processing power, storage space, and battery life. Whereas cloud and fog computing help to relieve computational and storage constraints, energy limitations remain a severe impediment to long-term autonomous operation. Among the threats that exploit this weakness, the Denial-of-Sleep (DoSl) attack is particularly problematic because it prevents nodes from entering low-power states, leading to battery depletion and degraded network performance. This research investigates machine-learning (ML) and deep-learning (DL) methods for identifying such energy-wasting behaviors to protect IoT energy resources. A dataset was generated in a simulated IoT environment under multiple DoSl attack conditions to validate the proposed approach. Several ML and DL models were trained and tested on this data to discover distinctive power-consumption patterns related to the attacks. The experimental results confirm that the proposed models can effectively detect anomalous behaviors associated with DoSl activity, demonstrating their potential for energy-aware threat detection in IoT networks. Specifically, the Random Forest and Decision Tree classifiers achieved accuracies of 98.57% and 97.86%, respectively, on the held-out 25% test set, while the Long Short-Term Memory (LSTM) model reached 97.92% accuracy under a chronological split, confirming effective temporal generalization. All evaluations were conducted in a simulated environment, and the paper also outlines potential pathways for future physical testbed deployment. Full article

Accurate day-ahead solar forecasting is essential for grid stability and energy planning. This study introduces a specialized forecasting framework that enhances accuracy by training models on specific day-to-day sky condition transitions. The framework employs a dual-attention transformer model, which captures complex temporal and feature-wise relationships, using a dataset of approximately 5000 daily sequences from three sites in Mongolia (2018–2024). Our core contribution is a specialized training strategy where the dataset is first classified into nine distinct classes based on the sky condition transition from the previous day to the forecast day, such as ‘Clear’ to ‘Partly cloudy’. A dedicated transformer model is then trained for each transitional state, enabling it to become an expert on that specific weather dynamic. This specialized framework is benchmarked against a naive persistence model, a standard, generalized transformer trained on all data and a ‘cluster-then-forecast’ approach. Results show the proposed approach achieves superior performance improvement compared to baseline models (p < 0.001) across all error metrics, demonstrating the value of modeling inter-day weather dynamics. Furthermore, the framework is extended to probabilistic forecasting using quantile regression to generate 80% prediction intervals, providing crucial uncertainty information for operational decision-making in power grids. Full article

Background: Volleyball requires explosive jumps, agility, and upper and lower limb coordination. Dynamic stretching (DS) and core stability (CS) protocols are often used separately in training sessions, but little is known about their combined effects on the performance in adolescent players. This study aimed to investigate the impact of a 12-week integrated DS and CS program (StretCor), in addition to standard training, on physical performance in U-16 volleyball players. Methods: Twenty-one volunteer players (15.1 ± 0.6 years) were randomly assigned to the Intervention Group (IG; n = 12) or Control Group (CG; n = 9). IG performed the StretCor protocol four times a week for twelve weeks in addition to standard volleyball training; CG continued standard volleyball training. Physical performance assessment included Countermovement Jump (CMJ), Vertec jump with run-up, isometric shoulder strength (ASH-I), dynamic balance (mSEBT), and agility (t-test) tests. Results: Significant group × time interactions (p < 0.05, η² ranged: 0.20–0.90) were found for CMJ height and peak power, Vertec jump, ASH-I, mSEBT scores, and t-test performance. Post hoc analyses showed improvements in IG for CMJ height (+16.5%), Vertec jump (+10.2%), shoulder strength (+11–14%), balance across directions (+8–12%), and agility (−5.7% t-test time). No significant changes were observed in CG. Conclusions: The present study suggests that a 12 weeks of StretCor protocol training improves jump performance, agility, dynamic balance, and upper limb strength in U-16 volleyball players. These findings also support that StretCor protocol may be beneficial for the performance when incorporated into regular training programs for adolescent athletes. Full article

Music is a near-universal anthropological and sensory phenomenon that engages distributed brain networks and peripheral physiological systems to shape emotion, cognition, sociality, and bodily regulation. Evidence from electrophysiology, neuroimaging, endocrinology, randomized controlled trials, and longitudinal training studies indicates that both receptive and active musical experiences produce experience-dependent neural and systemic adaptations. These include entrainment of neural oscillations, modulation of predictive and reward signaling, autonomic and neuroendocrine changes, and long-term structural connectivity alterations that support affect regulation, cognition, social functioning, motor control, sleep, and resilience to neuropsychiatric illness. This narrative review integrates mechanistic domains with clinical outcomes across major conditions, such as depression, anxiety, schizophrenia, dementia, and selected neurodevelopmental disorders, by mapping acoustic and procedural parameters onto plausible biological pathways. We summarize how tempo, beat regularity, timbre and spectral content, predictability, active versus passive engagement, social context, dose, and timing influence neural entrainment, synaptic and network plasticity, reward and prediction-error dynamics, autonomic balance, and immune/endocrine mediators. For each condition, we synthesize randomized and observational findings and explicitly link observed improvements to mechanistic pathways. We identify methodological limitations, including heterogeneous interventions, small and biased samples, sparse longitudinal imaging and standardized physiological endpoints, and inconsistent acoustic reporting, and translate these into recommendations for translational trials: harmonized acoustic reporting, pre-specified mechanistic endpoints (neuroimaging, autonomic, neuroendocrine, immune markers), adequately powered randomized designs with active controls, and long-term follow-up. Contextual moderators including music education, socioeconomic and cultural factors, sport, sleep, and ritual practices are emphasized as critical determinants of implementation and effectiveness. Full article

Data augmentation is one of the effective solutions to improve the performance of machine learning models in general and deep learning in particular. Data augmentation techniques bring different effects to each model, but very few studies have considered this issue. This study investigated the effect of five distinct data augmentation strategies on a custom-built Convolutional Neural Network (CNN) and nine pre-trained CNN models for crack detection. All ten models were initially trained on a reference dataset of unaugmented images, followed by separate experiments using the augmented datasets. The results show that the pre-trained models, especially VGG-16, EfficientNet-B7, Xception, DenseNet-201, and EfficientNet-B0, consistently achieved greater than 98% in accuracy across all augmentation techniques. Meanwhile, the custom-built CNN was very sensitive to illumination changes and noise. Image rotation and cropping have minimal negative impact and sometimes improve performance. The findings demonstrate that combining data augmentation with state-of-the-art pre-trained models offers a powerful and efficient alternative to the reliance on large-scale datasets for accurate crack detection using CNNs. Full article

While prior research has examined the role of smart technologies (e.g., IoT and AI) in sustainability, the combined influence of IoT, AI, and organizational capabilities on hotel sustainable performance, particularly through the mediating roles of data-driven decision-making and innovation capability, remains underexplored. This study investigates how the integration of smart technologies, specifically the Internet of Things (IoT) and artificial intelligence (AI), as well as dynamic managerial capabilities focusing on data-driven decision-making (DDM) and innovation capability (IC), enhances hotel sustainable performance (HSP) within the context of Saudi Arabia’s hospitality sector. Grounded in the Resource-Based View (RBV) and Dynamic Capabilities Theory (DCT), the research develops and tests a conceptual model that explores both the mediating roles of DDM and IC in the link between IoT and HSP and the moderating role of AI application in the relationships between IoT and DDM, IC, and HSP. Using data collected from 312 managers of four- and five-star hotels across Saudi Arabia, the study employed Partial Least Squares Structural Equation Modeling (PLS-SEM) to examine the hypothesized relationships. The results reveal that IoT has a significant positive effect on HSP, DDM, and IC. Further, the IoT-HSP relationship is partially mediated by both DDM and IC. Furthermore, AI significantly strengthens the relationships between IoT and DDM, IoT and IC, and IoT and HSP, highlighting AI’s crucial role as an enabler of digital transformation and sustainability. The findings extend the RBV and DCT by demonstrating how technological resources, when combined with dynamic managerial capabilities, lead to superior sustainability outcomes. Practically, the study emphasizes that hotels must pair digital adoption with employee training, innovation culture, and AI-powered analytics to enhance HSP. Full article

With advances in computing power and deep learning, image-based pose estimation has become a viable tool for quantitative motion analysis. Compared to sensor-based systems, vision-based approaches are cost-effective, portable, and easy to deploy. However, when applied to golf swings, conventional similarity measures often fail to match expert perception, as they rely on static, frame-wise posture comparisons and require strict temporal alignment. We propose a Dynamic Motion Similarity Measurement (DMSM) framework that segments a swing into seven canonical phases—address, takeaway, half, top, impact, release, and finish—and evaluates the dynamic trajectories of joint keypoints within each phase. Unlike traditional DTW- or frame-based methods, our approach integrates continuous motion trajectories and normalizes joint coordinates to account for player body scale differences. Motion data are interpolated to improve temporal resolution, and numerical integration quantifies path differences, capturing motion flow rather than isolated postures. Quantitative experiments on side-view swing datasets show that DMSM yields stronger discrimination between same- and different-player pairs (phase-averaged separation: 0.092 vs. 0.090 for the DTW + cosine baseline) and achieves a clear biomechanical distinction in spine-angle trajectories (Δ = 38.68). Statistical analysis (paired t-test) confirmed that the improvement was significant (p < 0.05), and coach evaluations supported perceptual alignment. Although DMSM introduces a minor computational overhead (≈169 ms), it delivers more reliable phase-wise feedback and biomechanically interpretable motion analysis. This framework offers a practical foundation for AI-based golf swing analysis and real-time feedback systems in sports training, demonstrating improved perceptual consistency, biomechanical interpretability, and computational feasibility. Full article

Professional skills training and academic success are key challenges for contemporary educational systems, particularly within higher education. The labour market increasingly demands well-prepared graduates with specific competencies that are still insufficiently embedded in university curricula. In this context, acquiring new professional skills becomes a decisive factor for students’ employability and competitiveness. At the same time, academic success remains a crucial indicator of educational quality, and its improvement is an urgent priority for universities. In response to these demands, our study evaluates cognitive-emotional competencies—emotional intelligence, creativity, and subjective well-being—in a sample of 300 university students from the Community of Madrid (Spain), analysing their influence on academic success with the aim of enhancing it. A non-experimental, cross-sectional research design was employed, using standardised self-report measures (TMMS-24, CREA, SHS, OHI, SLS, and OLS), innovative data mining algorithms (Random Forest and decision trees), and binary logistic regression techniques. The results highlight the importance of creativity, life satisfaction, and emotional attention in predicting academic success, with creativity showing the strongest discriminative power among the variables studied. These findings reinforce the need to integrate emotional and creative development into university curricula, promoting competency-based educational models that enhance training quality and students’ academic outcomes. Full article

Sepsis is a life-threatening condition driven by a dysregulated immune response, leading to systemic inflammation and multi-organ failure. Among the key molecular regulators, S100A8/A9 has emerged as a critical damage-associated molecular pattern (DAMP) protein, amplifying pro-inflammatory signaling via the Toll-like receptor 4 (TLR4) and receptor for advanced glycation end products (RAGE) pathways. Elevated S100A8/A9 levels correlate with disease severity, making it a promising biomarker and therapeutic target. To unravel the role of S100A8/A9 in sepsis, we integrate scRNA-seq and RNA-seq approaches. scRNA-seq enables cell-type-specific resolution of immune responses, uncovering cellular heterogeneity, state transitions, and inflammatory pathways at the single-cell level. In contrast, RNA-seq provides a comprehensive view of global transcriptomic alterations, allowing robust statistical analysis of differentially expressed genes. The integration of both approaches enables precise deconvolution of immune cell contributions, validation of cell-specific markers, and identification of potential therapeutic targets. Our findings highlight the S100A8/A9-driven inflammatory cascade, its impact on immune cell interactions, and its potential as a diagnostic and prognostic biomarker in sepsis. Eight protein targets resulted from the integrative transcriptomics studies (corresponding to S100A8, S100A9, S100A6, NAMPT, FTH1, B2M, KLF6 and SRGN) have been used to predict interaction affinities with 2958 ChEMBL approved drugs, by using a pre-trained AI models (PLAPT) in order to point directions on drug repurposing in sepsis. The strongest predicted interactions have been confirmed with molecular docking and molecular dynamics analysis. This study underscores the power of combining high-throughput transcriptomics to advance our understanding of sepsis pathophysiology and develop precision medicine strategies. Full article

The rapid proliferation of photovoltaic (PV) generation has transformed conventional distribution systems, resulting in frequent reverse power flow (RPF) and associated overvoltage issues. This paper presents a deep reinforcement learning (DRL)-based topology control method to autonomously mitigate RPF and voltage violations. A novel multi-discrete Maskable Proximal Policy Optimization (MPPO) algorithm is proposed, combining topology-aware action masking with a multi-discrete action representation to ensure constraint satisfaction and enhance training stability. The approach efficiently explores the feasible switching space while maintaining network radiality, load connectivity, and power flow solvability. Extensive case studies based on one year of operational data from a practical distribution system show that the proposed agent achieves an average RPF reduction of 24.3% across the test cases and restores normal voltage conditions in about 65% of scenarios, while satisfying other operational constraints. The results confirm that the proposed method provides a scalable, data-driven solution for topology reconfiguration in PV-rich distribution networks. Full article

Distributed photovoltaic power generation is volatile and intermittent, and its power generation is usually difficult to accurately predict. Previous studies have focused on physical or mathematical modeling methods, and it is difficult to grasp the complexity and variability of historical data, and the prediction accuracy is limited. To address these challenges, this paper proposes Solar-LLM, a novel prediction framework that adapts a pre-trained Large Language Model (LLM) for time-series forecasting. By freezing the core LLM and reprogramming only its input and output layers, Solar-LLM efficiently translates numerical time-series data into a format the model can understand. This approach leverages the LLM’s inherent ability to capture long-term dependencies and complex patterns, enabling effective learning even from limited data. Experiments conducted on a dataset from five photovoltaic power stations show that Solar-LLM significantly outperforms baseline models, proving it to be a highly effective and feasible solution for distributed PV power prediction. Full article