Search Results (5,742)

This study systematically reviews the application of machine learning (ML) in emergency resource allocation across Europe, with the aim of synthesizing current evidence and identifying future research directions. A systematic literature review (SLR) was conducted following PRISMA guidelines. Data were collected from major academic databases (2018–2025) using predefined inclusion and exclusion criteria. A total of 52 relevant studies were analyzed through qualitative thematic synthesis. The review finds that ML significantly enhances predictive analytics, enabling accurate forecasting of emergency demand and proactive resource allocation. ML-driven optimization improves ambulance dispatch, hospital resource management, and logistics efficiency, while real-time decision support systems strengthen situational awareness and coordination. However, challenges persist, including data quality issues, system fragmentation, ethical concerns (bias, transparency), and limited interoperability across European systems. ML has transformative potential in shifting emergency resource allocation from reactive to data-driven, predictive systems. Its effectiveness, however, depends on robust data infrastructure, ethical governance, and system integration. The study recommends strengthening data systems, adopting hybrid ML-optimization models, enhancing ethical frameworks, investing in human capacity, and promoting cross-border collaboration. Full article

Layered cloud cover (LCC) describes the vertical distribution of cloud occurrence and is a key variable for assessing the radiation budget of the Earth-atmosphere system. However, ground-based radars have limited spatial coverage, while existing satellite cloud-cover products rarely provide both spatiotemporal continuity and high accuracy. Because nighttime satellite observations lack visible-channel information, conventional passive satellite remote sensing remains limited in providing day-night continuous LCC retrievals. In this study, we propose an infrared-based framework for retrieving large-scale day-night LCC from geostationary satellite observations. The framework first resolves cloud vertical structure using a hybrid machine learning and physical algorithm for day-night cloud-base height (CBH) retrieval, and then derives cloud cover in different vertical layers. Validation against active measurements from spaceborne and ground-based cloud radar demonstrates that the satellite-retrieved LCC captures cloud vertical distributions and their diurnal variations. The cloud-layer identification accuracies reach 76.3% and 77.9% for daytime and nighttime, respectively, with corresponding Cohen’s kappa coefficients of 0.66 and 0.68. The primary source of algorithmic uncertainty is the low precision of low-cloud identification, which is constrained by objective factors and physical characteristics. The retrieved annual mean LCC fields reproduce major climatological features, including enhanced high and deep convective clouds over the tropical western Pacific and dominant low-cloud occurrence over the mid-latitude oceans. A case study of Typhoon Doksuri further shows that the 10 min LCC retrievals capture the vertical evolution of the typhoon cloud system during intensification, eyewall structural adjustment, landfall, and post-landfall decay. These results indicate that the proposed infrared-based retrieval framework provides a promising basis for constructing large-scale day-night LCC datasets and can support cloud-radiation studies, climate-model evaluation, and weather monitoring. Full article

This work aims to study and improve the long-term stability of conductimetric biosensors for urea detection in clinical and environmental samples, which are fundamentally limited by complex thermal and temporal drifts due to temperature-sensitive enzyme kinetics, variations in ionic mobility, and the progressive degradation of the sensing layer. The biosensor targets the urea concentration range 0.01–30 mM, validated against experimental data and covering the clinically relevant range for blood urea detection (2.5–7.5 mM), urine (20–40 mM), and environmental monitoring applications. Conventional calibration techniques, such as the conventional calibration method (based on reference measurements), and purely deterministic correction methods, such as deterministic methods (based on known fixed equations), often prove insufficient because they struggle to capture the non-stationary and inherently stochastic nature of these drifts. In this work, we propose an original hybrid probabilistic framework that synergistically combines machine learning and Bayesian inference for robust adaptive drift compensation. A Random Forest model is first implemented to model the deterministic nonlinear relationships between environmental parameters (temperature, pH, CO₂ concentration) and the sensor response. The residual temporal drift is then explicitly modeled as a non-stationary latent stochastic process using Bayesian inference based on a Gaussian process. This approach allows continuous online model updating, real-time uncertainty quantification, and automatic detection of anomalies. The models were trained and validated on a large dataset obtained from multiphysics simulations carried out in COMSOL Multiphysics 5.6. These simulations incorporated enzymatic reactions, thermal effects, and chemical dynamics taking place inside the sensor. Experimental results show that the hybrid approach substantially enhances sensor performance, lowering the root mean square error (RMSE) to below 0.8 μS/cm (corresponding to less than 0.5% of the full-scale response) over a wide temperature range (15–45 °C) and across extended operating periods. This represents a clear improvement over conventional compensation method. By merging the predictive power of ensemble learning with a probabilistic Bayesian model of dynamic drift, this study introduces a fresh perspective on the design of intelligent, self-adaptive, and drift-resistant conductimetric biosensors. The proposed framework holds strong potential for reliable, long-term autonomous operation in urea reliable, long-term autonomous operation in urea monitoring across biomedical diagnostics (kidney/liver function assessment) and environmental surveillance (water eutrophication prevention). Full article

Street-level concentrations of nitrogen dioxide (NO₂) and fine particulate matter (PM_2.5) pose serious public health risks in European cities, yet accurate multi-year prediction at traffic-dominated sites remains challenging. This study applies XGBoost (XGB) and Random Forest (RF) to predict hourly NO₂ and daily PM_2.5 at two street monitoring sites in Copenhagen, Denmark, trained on 17 years of observational data and evaluated on two independent years. Three-dimensional variational assimilation (3D-Var) and the Extended Kalman Filter (EKF) are then applied as post-processing corrections to the ML predictions using co-located observations. XGB achieved RMSE values of 9.5 and 7.4 µg/m³ for HCAB and JGTV NO₂, respectively, in the 2018 test year. Both DA methods improved substantially on the ML baseline, with 3D-Var reducing NO₂ RMSE by up to 57% and spike event RMSE by up to 51%. EKF achieved near-complete elimination of systematic bias across all configurations. The framework is computationally lightweight and can be applied to any deterministic model prediction at a monitoring station, including outputs from physics- and chemistry-based dispersion models. Overall, the findings show a practical way to improve street-level air quality prediction, with direct relevance for operational forecasting and public health protection. Full article

Low-resource hate speech detection remains a challenging task for Indonesian social media due to limited labeled annotations, highly informal linguistic expressions, and substantial lexical variability. Under such conditions, purely supervised transformer models often suffer from unstable semantic generalization, while conventional pseudo-labeling methods are vulnerable to noisy unlabeled sample propagation. To address these limitations, this study proposes a hybrid feature-enhanced IndoBERT framework integrated with a controlled semi-supervised learning strategy. The proposed model combines contextual IndoBERT embeddings with abusive lexicon cues, handcrafted linguistic indicators, and TF-IDF–SVD statistical representations through a lightweight concatenation–projection feature fusion mechanism, while unlabeled data are incorporated via adaptive confidence thresholding and class-balanced pseudo-label selection to improve pseudo-label reliability. Extensive experiments were conducted under realistic low-resource supervision settings using only 5%, 10%, and 20% labeled data, and the proposed framework was systematically compared against representative baselines, including sparse lexical machine learning models, shallow neural architectures, multilingual transformers, IndoBERTweet, naive pseudo-labeling, and LLM-based prompting. The results show that model effectiveness is strongly supervision-dependent. Under the most extreme low-resource setting, compact statistical augmentation provides the most stable complementary signal, whereas under moderate low-resource supervision, the full hybrid representation combined with controlled semi-supervised learning yields the strongest and most consistent gains. The proposed Hybrid IndoBERT + controlled SSL framework outperforms all baselines at the 20% labeled setting, reaching an accuracy of 0.8654, Macro-F1 of 0.8633, and ROC-AUC of 0.9334. Additional analyses of pseudo-label reliability, calibration behavior, computational efficiency, and qualitative error patterns further show that the proposed framework improves low-resource robustness while maintaining comparable inference-time efficiency. These findings demonstrate that low-resource hate speech detection benefits most from the staged integration of contextual semantic modeling, interpretable linguistic cues, global lexical–statistical structure, and carefully regulated unlabeled data exploitation. Additional experiments using GPT-4o-mini and Llama-3.1-8B further demonstrate that the proposed framework remains competitive against general-purpose large language model prompting approaches under low-resource Indonesian hate speech detection scenarios. The proposed framework provides a practical and reproducible direction for hate speech detection in annotation-constrained social media environments. Full article

Actual evapotranspiration (ETa) is a key component of the hydrological cycle, and its partitioning into soil evaporation (Es) and vegetation transpiration (Ec) is essential for understanding hydrological processes. Focusing on the Yiluo River Basin during 1960–2020, this study developed a hybrid framework combining the physically based Penman–Monteith–Leuning (PML) model with machine learning to dynamically parameterize the soil evaporation coefficient f. ERA5-Land reanalysis data were used to drive the model, while the Pettitt change-point test and ridge regression were applied to identify potential change points and quantify driving factors. The results show that the framework improved the agreement of Es simulations with the GLEAM-derived reference product (R² and NSE > 0.8) and reduced the difference in ETa estimates by approximately 10% within the product-constrained modeling framework. ETa exhibited a significant upward trend (0.28 mm·yr⁻¹) with a potential change point around 2004, while its components responded earlier, with Ec and Es changing in 1994 and 2002. Ec dominated ETa, accounting for about 70% of the total. Net radiation, temperature and leaf area index were primary controls, while increasing vapor pressure deficit, together with changes in relative humidity and precipitation, jointly regulated the identified shifts. These findings provide a process-based understanding of ETa dynamics and improve the representation of ETa components in hydrological modeling. Full article

Background/Objectives: The precise identification of Alzheimer’s disease (AD) stages through clinical data is crucial for early diagnosis and suitable therapy. This classification remains troublesome due to overlap in cognitive profiles across different phases of illness progression. This study presents a comprehensive and advanced diagnostic system, termed FLAME, featuring an enhanced federated learning architecture for privacy-preserving multi-institutional implementation. It provides a systematic review of machine learning (ML) and deep learning (DL) models for the classification of five stages of Alzheimer’s disease (AD). The models include cognitively normal (CN), subjective memory complaints (SMC), early mild cognitive impairment (EMCI), late mild cognitive impairment (LMCI), and Alzheimer’s disease (AD). Methods: Sixteen traditional machine learning models and eleven deep learning architectures—including FT-Transformer and NODE—were evaluated using a structured clinical dataset comprising 362 features. A hybrid ensemble was created at the probability level by combining the two top-performing models, LightGBM and a five-layer DNN. The weights of this ensemble were automatically optimised using a Genetic Algorithm (GA) with Macro-F1 as the fitness criterion, confirmed stable across 30 independent runs ($w^{★}=0.5024\pm 0.0001$). A federated learning architecture was then established, deploying the DNN across non-IID clients while keeping LightGBM centralised. We examine four distinct aggregation algorithms: FedAvg, FedProx, FedNova, and SCAFFOLD. Results: Among all deep learning architectures, FT-Transformer achieved the highest standalone performance (accuracy = 0.7810, $\kappa$ = 0.7081). The five-layer deep neural network (DNN) was selected as the DL representative for the hybrid ensemble. LightGBM attained superior machine learning performance (accuracy = 0.8156, $\kappa$ = 0.7537), confirmed deterministic across 10 seeds. The LightGBM vs. XGBoost difference is not statistically significant (McNemar $p=0.4227$). The GA-optimised hybrid ensemble (w = 0.685) surpassed both individual baselines across all evaluation metrics. The FedNova hybrid design achieved superior overall performance in federated configurations, surpassing all centralised arrangements in accuracy (accuracy = 0.8213, $\kappa$ 0.7614). Conclusions: Evolutionary ensemble optimisation combined with federated learning provides a robust, scalable, and privacy-preserving solution for AD stage classification, offering a clinically viable framework for real-world multi-institutional decision-support systems. However, the AD class remains severely under-recalled across all configurations (F1 ≤ 0.21), identifying this as the primary open challenge for clinical translation. Full article

Edge artificial intelligence (Edge AI) requires real-time inference under stringent constraints on computation, memory, energy, and connectivity. Although training can be offloaded to servers, efficient, high-capacity inference and rapid on-device adaptation remain central challenges. Cloud-based inference offers substantial computational power but depends on connectivity, latency, privacy, and reliability conditions that edge deployments cannot always guarantee. Classical model-compression methods—including quantization, pruning, distillation, and neural architecture search—have extended the feasibility of on-device inference, yet they leave largely unchanged the fundamental cost of the linear-algebraic, sampling, and optimization primitives that dominate modern deep learning. Quantum computing has therefore been proposed as a complementary accelerator for selected AI workloads, with theoretical advantages in linear systems, singular value decomposition, sampling, kernel evaluation, and optimization. This review surveys the emerging field of quantum-accelerated AI for edge systems under a hybrid architectural premise: edge devices remain classical, while quantum processors operate as remote, cloud, MEC, or near-edge accelerators. We synthesize advances across quantum learning models, hybrid optimization methods, hardware and deployment architectures, and quantum-inspired approaches suitable for constrained devices. We also assess the practical barriers that currently separate asymptotic quantum advantage from deployable edge intelligence, including data loading, measurement overhead, noise, latency, and benchmarking gaps. Finally, we outline a staged research roadmap from near-term hybrid workflows to fault-tolerant and integrated quantum-edge architectures. Full article

This study investigates the electronic and nonlinear optical properties of coupled GaN/AlN quantum dots using a numerical approach based on coordinate transformation combined with the finite difference method (FDM). The Schrödinger equation is solved to determine the electronic energy levels and wave functions of the system, which are subsequently used to evaluate the nonlinear optical rectification (NOR) response. Since numerical simulations become computationally expensive for large quantum dot systems, several regression-based models, including Polynomial Regression, Ridge Regression, LASSO, and Elastic Net, are trained on high-fidelity numerical data. These models learn the relationship between structural parameters and the resulting electronic and optical properties, enabling fast and reliable predictions for larger quantum dot configurations. The predictive performance of the ML models is assessed by comparing their results with the numerical simulations, showing excellent agreement while significantly reducing computational effort. The proposed hybrid physics–machine learning framework therefore provides an efficient and reliable approach for predicting the electronic and nonlinear optical behavior of coupled GaN/AlN quantum dots. Full article

Artificial intelligence (AI) and machine learning (ML) are increasingly important in financial technology (FinTech) applications involving large datasets, uncertainty, and complex decision-making. First, this paper presents a review of AI- and ML-based approaches in FinTech from 2010 to 2025, with particular emphasis on the relationship between predictive analytics and optimization-based decision-making. The review identifies two major research streams: (i) predictive AI/ML models for financial forecasting, stock price prediction, risk management, and fraud detection and (ii) optimization approaches for constrained financial decision problems, including portfolio optimization, asset–liability management, and risk-based decision-making. These two streams have largely evolved independently, which creates challenges in real financial environments, where uncertainty in predictions directly affects decision quality. Secondly, the paper also provides a decision-oriented perspective on how AI/ML-based predictions can support optimization under uncertainty and practical financial constraints. It highlights the role of uncertainty-aware optimization, simulation-based methods, and hybrid approaches such as simheuristics in improving the robustness of financial decision-making. Finally, the paper identifies open research directions toward integrated financial decision-support frameworks that combine predictive analytics, optimization, and simulation to address dynamic and uncertain FinTech environments. Full article

Objective Image Quality Assessment (IQA) is a fundamental pillar of computer vision, essential for optimizing tasks ranging from supervised machine learning to real-time video streaming. While IQA aims to quantify image degradation caused by noise and artifacts, a persistent gap remains between technical objective measurements and subjective human perception. Objective IQA has advanced significantly through full-reference (FR) metrics designed to approximate human judgment. Standard measures such as the peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and root mean square error (RMSE) provide established benchmarks; however, they frequently fail to capture nuanced human visual preferences, often penalizing perceptually insignificant shifts or favoring overly smoothed images. Conversely, modern deep-learning metrics like LPIPS offer better perceptual alignment but remain computationally prohibitive for real-time, resource-constrained environments. This paper addresses these challenges through a dual-purpose approach. First, it provides a comprehensive survey and tutorial of the IQA landscape, offering self-contained mathematical derivations of classical error sensitivity measures, including MSE, RMSE, MAE, Euclidean distance, RMSLE, and Huber loss, as well as artificial neural network (ANN) approaches. This foundational review ensures a rigorous understanding of the field’s mathematical evolution. We introduce the Adaptive Contrast-Weighted Structural Similarity (ACSSIM) framework. ACSSIM is a lightweight hybrid metric that enhances classical FR-IQA by incorporating local weighting derived from human visual system (HVS) properties. Specifically, it targets Weber’s Law-based contrast and entropy, which are key elements of our hybrid quality assessment logic and key components of non-reference image quality metrics. Extensive numerical experiments on the TID2013 and KADID-10k benchmark show that ACSSIM improves correlation with human subjective judgments compared with the baseline PSNR and SSIM. Our results confirm that ACSSIM maintains low computational overhead, bridging the gap between efficiency and accuracy for practical deployment. We made our code publicly available to facilitate future research in efficient perceptual modeling. Full article

Hybrid additive manufacturing (HAM) integrates additive and subtractive processes within a unified production system, combining the geometric flexibility and material efficiency of additive manufacturing with the dimensional accuracy and surface quality of conventional machining. This review provides a comprehensive analysis of HAM technologies through a proposed four-criterion classification framework encompassing process integration strategy, additive manufacturing process type, machine architecture, and application domain. DED-based, PBF-based, and polymer-based hybrid systems are examined alongside integrated hybrid machines, retrofit solutions, and robotic architectures. A comparative analysis of representative commercial platforms evaluates build envelope, integration strategy, and monitoring capability. Documented performance outcomes across aerospace, automotive, energy, and biomedical sectors confirm substantial improvements in surface quality, fatigue performance, dimensional accuracy, and material efficiency relative to conventional manufacturing routes. Current limitations are critically assessed across technical, process integration, and economic dimensions, and a structured near-to-long-term research roadmap is proposed, prioritising in-process sensing and toolpath standardisation, digital twin-based adaptive process planning, and ultimately autonomous hybrid manufacturing cells with lifecycle certification. These findings position HAM as a central enabling technology for intelligent, flexible, and sustainable production within Industry 4.0 and Industry 5.0 paradigms. Full article

To address the challenge of maintenance decision-making for critical components in electro-hydraulic servo material fatigue testing machine, characterized by weak state observability and difficulty in degradation prediction, a multi-component joint maintenance decision-making method based on multi-head deep reinforcement learning is proposed. Considering the heterogeneity of the degradation mechanisms and observation methods for the four components—bearing beam, fixture, main machine sensors, and hydraulic oil tank—a continuous-discrete hybrid state Markov decision process (HS-MDP) is constructed. To account for differences in maintenance strategies across components, a differentiated discrete action space for each component is designed, and engineering feasibility constraints are explicitly integrated into the policy through action masking. A data-quality loss term, determined by the degradation level of the sensors, is introduced into the reward function to align the optimization objective with the metrological properties of the fatigue testing machine. Based on the Branching Dueling DQN framework, a Q-network structure is constructed, incorporating a shared encoder, an inter-component attention mechanism, and multi-head branched outputs. Taking a 100 kN electro-hydraulic servo fatigue testing machine as a case study, comparisons with baseline strategies such as periodic maintenance, threshold-based condition-based maintenance (CBM), independent DQN, and PPO indicate that the proposed method reduces the average annual total cost by 60.3% compared to periodic maintenance and by 42.6% compared to threshold-based CBM. The number of failures decreases from 9.8 times/year to 1.4 times/year, while data efficiency increases from 82.1% to 96.2%. Ablation experiments and robustness tests further verify the critical contributions of three key design elements: action masking, inter-component attention, and data-quality loss. Full article

Full-physics reservoir simulation for CO₂ storage becomes computationally expensive when many operational schedules must be screened, motivating machine-learning (ML) surrogates that reduce simulation burden while preserving the essential physics-driven response. We propose an activity-based partitioning methodology that produces an interpretable applicability map, identifying regions where surrogate substitution is expected to be reliable and regions where highly active dynamics make it unsafe. In this work, we focus exclusively on the slow-varying region and develop proxy models for pressure and saturation time derivatives in that domain. The fast-varying region is intentionally excluded, and no fully coupled hybrid simulator is claimed at this stage. The partition is constructed from temporal changes in derivative signals and aggregated across multiple schedules to obtain a conservative, scenario-robust delineation. For slow cells, local stencil-based neural proxies leverage overlapping time windows and features describing the local state, schedule forcing, and injector influence. Because saturation derivatives in the slow region are strongly zero-inflated, with many cells remaining outside the advancing CO₂ plume for long periods, a two-stage strategy is adopted: first detecting whether meaningful change occurs and then predicting the derivative magnitude only when active, with additional smoothing to suppress near-zero artifacts. The framework also supports selective surrogate deployment over user-selected time windows. The objective is therefore to establish a conservative zone of applicability for derivative-based ML updates, rather than to demonstrate full simulator replacement or end-to-end coupled acceleration. In the case study, 5914 of the 8243 grid blocks evaluated by the proxy workflow were classified as slow-varying, corresponding to 71.7% of the evaluated proxy-analysis domain. For the blind schedule, full-rollout pressure reconstruction produced mean absolute errors of 5.34, 3.69, and 2.80 psi over early, middle, and late time-window groups, respectively. In a future coupled implementation using the same partition, these 5914 cells could be advanced by the ML proxy, while the remaining dynamically active or unsupported cells would remain under full-physics treatment. This would reduce the full-physics active-cell count from 9212 to 3298 in the future coupled setting, although direct wall-clock acceleration remains to be quantified after simulator integration. Full article

Flood-induced scour remains one of the leading causes of bridge failure, yet the chain linking flood hazard to bridge decisions is still commonly treated as a sequence of disconnected tasks. This review examines that chain using uncertainty as a unifying interpretive framework, synthesizing the recent literature on non-stationary flood hazard assessment, bridge-scale hydraulics, scour processes and predictive models, scour monitoring, monitoring-informed forecasting, structural vulnerability, and risk-informed decision-making. The review synthesizes the state of the art across all these stages of the chain, highlighting how the dominant uncertainty changes along it: climate and hydrologic variability upstream; model-form, sediment, and parameter uncertainty in scour prediction; measurement noise and inverse-inference uncertainty in monitoring; and threshold and consequence uncertainty in closure, retrofit, and network-level decisions. Although major advances have been achieved in probabilistic modelling, machine learning, hybrid physics-informed methods, and multimodal sensing, most published frameworks still transfer deterministic outputs from one stage to the next. As a result, uncertainty is rarely propagated consistently to the decision level. The main value of this review lies in making the chain’s weak interfaces explicit, in showing how uncertainty propagation can serve as a unifying framework across otherwise disconnected literatures, and in identifying which methodological directions are most promising for connecting prediction, monitoring, and decision support into a coherent end-to-end probabilistic chain supporting climate-resilient bridge management. Full article