Search Results (2,916)

This study develops a multiregional forecasting framework for road traffic accidents in Ecuador, addressing a critical limitation in existing predictive approaches that rely predominantly on point error metrics without validating the statistical assumptions underlying forecast uncertainty. Although the analysis is conducted at the provincial level, the spatial dimension is used primarily for cross-regional comparison and risk classification rather than for explicit spatial interaction modeling. Using a dataset of 27,648 monthly observations covering all 24 provinces from 2014 to 2025, the study applies the Prophet model within a Design Science Research paradigm and a CRISP-DM implementation cycle. Separate provincial models are estimated with a 24-month forecasting horizon, and methodological rigor is ensured through systematic residual diagnostics using the Shapiro–Wilk test for normality and the Ljung–Box test for temporal independence. Empirical results indicate that the Prophet-based artifact outperforms a naïve seasonal benchmark in 70.8% of the provinces, demonstrating excellent predictive accuracy in structurally stable regions such as Tungurahua (MAPE = 10.9%). At the same time, the framework enables the identification of critical emerging risks in provinces such as Santo Domingo and Cotopaxi, where projected increases exceed 49% despite acceptable point forecasts. The findings confirm that point accuracy alone does not guarantee the validity of confidence intervals and that residual validation is essential for trustworthy uncertainty quantification. Overall, the proposed approach provides a robust foundation for a predictive surveillance system capable of supporting differentiated, evidence-based road safety policies in territorially heterogeneous contexts. Full article

In the context of the coordinated development of high-proportion renewable energy integration and alternating current/direct current (AC/DC) hybrid transmission, the sending-end power grid faces challenges such as decreased system strength, contracted stability boundaries, and difficulties in covering high-risk operating conditions. This paper proposes a new renewable energy scale impact analysis method that integrates “typical scenario construction-scale ladder comparison–prediction-driven time series injection” in response to the operational constraints of AC/DC transmission. In terms of method implementation, firstly, a two-layer typical scenario system is constructed under unified transmission constraints and fixed grid boundaries: A regular benchmark scenario covers the main operating range, and a set of high-risk scenarios near the boundaries is obtained through multi-objective intelligent search, which is then refined through clustering to form a computable stress-test scenario library. Here, the boundary scenarios are generated by a multi-objective search that simultaneously drives multiple key section load rates towards their limits, subject to AC power-flow feasibility and operational constraints, and the resulting Pareto candidates are reduced into a compact stress-test library by clustering. Secondly, a ladder scenario with increasing renewable energy scale is constructed, and cross-scale comparisons are carried out within the same scenario system to extract the scale effect and critical laws of key safety indicators. Finally, data resampling and Gated Recurrent Unit multi-step prediction are introduced to generate wind power output time series, enabling the temporal mapping of prediction results to scenario injection quantities, and constructing a closed-loop input interface of “prediction–scenario–grid indicators”. The results demonstrate that the proposed hierarchical framework, under unified AC/DC export constraints, can effectively construct a compact stress-test scenario library with enhanced boundary-risk coverage and can reveal how transient voltage security evolves across renewable expansion scales. By coupling boundary-oriented scenario construction, cross-scale comparable assessment, and forecasting-driven time series injection, the framework improves engineering interpretability and practical applicability compared with conventional scenario sampling/reduction workflows. For the forecasting module, the Gated Recurrent Unit (GRU) model achieves MAPE = 8.58% and RMSE = 104.32 kW on the test set, outperforming Linear Regression (LR)/Random Forest (RF)/Support Vector Regression (SVR) in multi-step ahead prediction. Full article

Handgrip strength (HGS) is widely used as a biomarker of muscle function and overall health in older adults. However, conventional analyses based on peak force values may overlook relevant temporal features of the HGS curve. This cross-sectional study proposes a novel methodological approach that examines the shape and variability of HGS(t) curves recorded from community-dwelling older adults. Functional principal component analysis (FPCA) was applied to assess the consistency of individual trials and the representativeness of mean curves. Statistical non-parametric mapping (SnPM) was then used to identify time regions showing significant differences between groups. Complementary analyses of discrete and derivative parameters, together with non-parametric comparisons based on the Hodges–Lehmann estimator and corresponding 95% confidence intervals, were conducted to quantify effect sizes. FPCA revealed high within-participant consistency, supporting the use of mean curves for group-level comparisons. SPM analyses indicated significant differences in the early force development phase. Importantly, this approach shows that sex differences are attributable to magnitude effects, with men generating higher forces and faster early rates of force development, and not to differences in the neuromuscular strategy of force production. Traditional discrete parameters partly captured these patterns but failed to reflect the full temporal dynamics. This methodological approach to the HGS curve may provide further insights into neuromuscular control mechanisms that cannot be truly captured by the minimalistic HGS discrete parameters. Full article

Objectives: While the previous literature has established that meteorological conditions are associated with COVID-19 mortality fluctuations, the relative effect of each of these highly correlated factors remains unclear. This study aims to conduct a comparative analysis to determine which of three main meteorological variables—Ambient Temperature, Ultraviolet (UV) Index, and Sunlight Duration—have the strongest negative association with COVID-19 mortality. The objective is to quantify and rank their impact over a 7-to-21-day biological exposure window. Methods: We conducted retrospective spatiotemporal analyses in the form of panel Poisson Distributed Lag Models (PDLMs) regression using daily data from 21 January 2020 to 10 January 2023, spanning 129 distinct geographical regions worldwide. To ensure a direct and fair comparison of effect sizes, all meteorological and environmental variables were Z-score standardized. We estimated three independent PDLMs—each focusing separately on UV Index, Ambient Temperature, and Sunlight Duration—with lags ranging from 7 to 21 days. These models controlled for overarching time trends and utilized a categorical variable to account for Region Fixed Effects modeling time-invariant regional health and socioeconomic determinants (e.g., obesity, age demographics, healthcare capacity). Furthermore, distributed lags of daily PM2.5 (air pollution) and relative humidity were explicitly included in each model as dynamic confounders. Results: The comparison of PDLM results reveals that the UV Index has the strongest negative association with COVID-19 mortality. A one standard deviation increase in the UV Index corresponds to a massive, highly significant cumulative reduction in deaths observed 1 to 3 weeks later (p < 0.001). Sunlight Duration is the second-strongest protective meteorological factor, whereas Ambient Temperature has the weakest effect. The distributed lags of particulate matter (PM2.5) and relative humidity were found to be statistically insignificant when modeled alongside the meteorological variables. Conclusions: After standardizing variables and controlling for dynamic environmental confounders like air pollution and humidity, the study findings provide robust empirical evidence that meteorological conditions have a strong significant association with COVID-19 mortality fluctuation with a temporal delay, overcoming the confounding effects of merely dry or clear-air conditions. Full article

Patterns of footfall counts in urban environments show regularity at various spatial and temporal scales. In this work, we study a lightweight hierarchical approach in which forecasts use four lagged higher-level aggregates as predictors trained with simple CPU-only models. For a fair comparison, the baseline is expanded to use a horizon-matched lag window, so that the variants have access to the same maximum lookback in time. The study uses hourly pedestrian counts from 13 sensors on two shopping streets in Newcastle upon Tyne, aggregated across spatial and temporal levels. Combined spatial and temporal aggregate predictors reduced forecast error by adding information from higher aggregation levels without changing the base learner. The best-performing configuration was SHTH+CP, which combines spatial and temporal parent features with a spatio-temporal cross-parent, and yielded an average pooled 4.3% improvement in RMSE and 3.5% in MAE, with the largest gains at 12 h directional counts, where RMSE decreased by 6.7% and MAE by 11.4%. Full article

Despite widespread corporate endorsement of the Sustainable Development Goals (SDGs), systematic evidence on how top management in emerging economies prioritizes and frames SDG-related issues over time remains limited. Existing studies are often based on manual or single-year analyses, restricting comparability, scalability, and longitudinal insight. This study examines how corporate managerial communication aligns with and emphasizes SDGs across sectors and over time in two major emerging economies, India and South Africa. Using an AI-driven natural language processing (NLP) pipeline, we analyse 2400 annual reports from 600 publicly listed companies covering the period 2020–2023. A fine-tuned SDG-BERT multi-label classification model is applied to extract and classify SDG-related content from top management communications, enabling sectoral, temporal, and cross-country comparison of SDG relevance. The results reveal a strong and persistent emphasis on SDG 12 (Responsible Consumption and Production) across both countries, alongside sector-specific variation and differing patterns of SDG diversity over time. South African firms exhibit greater variation in SDG emphasis across years, while Indian firms display more concentrated and stable SDG framing. Overall, the findings highlight systematic imbalances in SDG-related managerial communication and persistent underrepresentation of several social SDGs. The study contributes methodologically by demonstrating the value of validated AI-assisted longitudinal text analysis for large-scale SDG research and empirically by providing comparative insights into how corporate SDG narratives evolve in emerging market contexts. Full article

Modern, post-Scaliger expectations for constructing an absolute chronology out of ancient biblical narratives introduce a fallacy of assumed time that distorts the reading of these narratives. While absolute chronology undergirds historical-critical interpretation from Spinoza and Reimarus to twentieth-century scholarship, the more recent “temporal turn” in philosophy, historiography, and literary theory aligns with a renewed attention to narrative time and ancient temporal consciousness. Focusing on 2 Maccabees and the Gospel of John as historiographical narratives reveals how both texts configure events through relative temporal devices—such as temporal markers and temporal process verbs—rather than through absolute calendrical dating, even when coordinates appear in 2 Maccabees’ embedded letters. Building on this comparison allows for a dimensional model of time that respects these configurational strategies and avoids obscuring how these texts construct theological and historical meaning within their own narrative worlds. Full article

Background: In this study, we aimed to evaluate and compare the diagnostic performance of peripapillary retinal nerve fiber layer (RNFL) thickness, macular ganglion cell–inner plexiform layer (GCIPL) thickness, and GCIPL asymmetry parameters in differentiating healthy eyes from primary angle-closure glaucoma (PACG), primary open-angle glaucoma (POAG), and secondary open-angle glaucoma (SOAG). Methods: This retrospective study included 204 eyes of 204 patients categorized into four groups: healthy controls (n = 46), PACG (n = 53), POAG (n = 58), and SOAG (n = 47). All participants underwent spectral-domain optical coherence tomography (OCT). Peripapillary RNFL thickness, sectoral and average GCIPL thickness, and GCIPL-derived asymmetry ratios were analyzed. Diagnostic performance was assessed using receiver operating characteristic (ROC) analysis. Results: Diagnostic accuracy varied according to glaucoma subtype. In distinguishing POAG from healthy controls, the average RNFL thickness (area under the ROC curve [AUC] = 0.82) demonstrated the highest diagnostic performance, followed by the superotemporal, inferotemporal, and average GCIPL thickness parameters. In contrast, no parameter reached an AUC of ≥0.80 in the PACG or SOAG comparisons. GCIPL asymmetry ratios exhibited limited discriminative ability across most analyses. Subtype differentiation was modest; POAG versus SOAG comparisons yielded AUC values up to 0.66, whereas PACG versus SOAG comparisons demonstrated minimal discrimination (AUC range: 0.47–0.63). Conclusions: Peripapillary RNFL and localized temporal GCIPL thickness measurements provide the highest diagnostic accuracy for identifying POAG. Diagnostic performance is reduced in PACG and SOAG, and the OCT parameters show limited ability to differentiate between glaucoma subtypes. GCIPL asymmetry indices do not enhance diagnostic discrimination beyond direct thickness measurements. Full article

Candidemia and invasive candidiasis remain significant causes of late-onset sepsis and mortality in very-low-birth-weight infants, especially in low- and middle-income countries. This narrative review synthesizes studies published between 2008 and 2025 in Latin America, addressing epidemiology, species distribution, antifungal susceptibility patterns, risk factors, therapeutic approaches, and clinical outcomes, with international comparisons. Accordingly, we present a qualitative narrative synthesis (see Methods) rather than a formal year-over-year temporal trend quantification. Globally, five species predominate, namely Candida albicans, C. parapsilosis sensu lato (s.I.), Candida tropicalis, Nakaseomyces glabratus, and Pichia kudriavzevii, with a sustained increase in non-albicans species and growing resistance to fluconazole. In Latin America, the burden varies depending on the hospital setting; C. parapsilosis sensu lato (s.I.) predominates in NICUs, and Candidozyma auris has emerged, associated with nosocomial outbreaks and multidrug resistance. Factors such as extreme prematurity, prolonged catheter use, parenteral nutrition, and antibiotics are consistently associated with the risk of infection. Mortality remains high, influenced by diagnostic delays and species characteristics. Standardized microbiological surveillance, accurate identification, and strategies tailored to each clinical setting are required to improve outcomes in this vulnerable population. Full article

The proliferation of Internet of Things (IoT) devices has intensified the need for efficient real-time anomaly and intrusion detection, making the selection of an appropriate Complex Event Processing (CEP) engine a critical architectural decision for security-aware data pipelines. Python-based CEP frameworks offer compelling advantages through the seamless integration with data science and machine learning ecosystems; however, rigorous comparative evaluations of such frameworks under realistic IoT security workloads remain absent from the literature. This study presents the first systematic comparative evaluation of Faust and Streamz—two Python-native CEP engines representing fundamentally different architectural philosophies—specifically in the context of IoT network intrusion detection. Faust was selected for its actor-based stateful processing model with native Kafka integration and distributed table support, while Streamz was selected for its reactive, lightweight pipeline design targeting high-throughput stateless processing, making them representative of the two dominant paradigms in Python stream processing. Although both engines target different application niches, their performance characteristics under realistic CEP workloads have never been rigorously compared, leaving practitioners without empirical guidance. The primary evaluation employs an IoT network intrusion dataset comprising 583,485 events from 83 heterogeneous devices. To assess whether the observed performance characteristics are specific to this single dataset or generalize across different workload profiles, a secondary IoT-adjacent benchmark is included: the PaySim financial transaction dataset (6.4 million records), selected because its event schema, fraud-pattern temporal structure, and volume differ substantially from the intrusion dataset, providing a stress test for cross-workload robustness rather than a claim of domain equivalence. We acknowledge the reviewer’s valid point that a second IoT-specific intrusion dataset (such as TON_IoT or Bot-IoT) would constitute a more directly comparable validation; this is identified as a priority for future work. The load levels used in scalability experiments (up to 5000 events per second) intentionally exceed the dataset’s natural rate to stress-test each engine’s architectural ceiling and identify saturation thresholds relevant to large-scale or multi-sensor IoT deployments. We conducted controlled experiments with comprehensive statistical analysis. Our results demonstrate that Streamz achieves superior throughput at 4450 events per second with 89% efficiency and minimal resource consumption (40 MB memory, 12 ms median latency), while Faust provides robust intrusion pattern detection with 93–98% accuracy and stable, predictable resource utilization (1.4% CPU standard deviation). A multi-framework comparison including Apache Kafka Streams and offline scikit-learn baselines confirms that Faust achieves detection quality competitive with JVM-based alternatives (Faust: 96.2%; Kafka Streams: 96.8%; absolute difference of 0.6 percentage points, not statistically significant at $p=0.318$) while retaining the Python ecosystem advantages. Statistical analysis confirms significant performance differences across all metrics ($p<0.001$, Cohen’s $d>0.8$). Critical scalability thresholds are identified: Streamz maintains efficiency above 95% up to 3500 events per second, while Faust degrades beyond 2500 events per second. These findings provide IoT security engineers and system architects with actionable, empirically grounded guidance for CEP engine selection, establish reproducible benchmarking methodology applicable to future Python-based stream processing evaluations, and advance theoretical understanding of the accuracy–throughput trade-off in stateful versus stateless Python CEP architectures. Full article

This paper considers a principle for constructing transverse differential protection of two parallel 6–35 kV power transmission lines on the supply side, previously proposed by the authors and protected by a patent. Its implementation makes it possible to identify the damaged line without using current and voltage transformers, by employing signals from reed switches installed near the corresponding phases of the lines. The principle is based on monitoring the difference in magnetic field inductions produced by the currents in these phases and determining the damaged line according to the sequence of reed switch actuations. Based on experimental studies of the temporal characteristics of reed switches and on simulation of magnetic fields around the line phases, methodologies for selecting protection operating parameters and for evaluating its sensitivity, taking into account the influence of errors and currents in the phases of both lines, have been developed; the field of application of the protection, the cascade zone, and the locations for installing the reed switches have been determined. It is shown that the calculated cascade zone of the proposed protection can be up to 20% shorter than that of both the traditional protection and a similar reed-switch-based protection. At the same time, in comparison with the latter, implementation of the considered principle requires three times fewer reed switches. The elimination of current and voltage transformers, as shown by the calculations, creates prerequisites for reducing the resource intensity of the protection. Full article

As a representative large shallow freshwater lake in China, Lake Taihu has suffered from persistent cyanobacterial blooms for a long time. Although intensive restoration actions have been carried out and caused visible improvements, the long-term evolution path and inner driving mechanisms of its ecosystem are still not fully made clear. Based on long-term monitoring data during 2011 to 2024, this study aims to characterize temporal dynamics of the aquatic environment, find out key drivers that shape community succession, and offer a scientific foundation for effective lake management. A series of data about hydrometeorological factors, physicochemical water quality indexes, and biological community data was analyzed by using the Mann–Kendall trend test, Pettitt change-point test, Redundancy Analysis, and correlation heatmaps. The results show that the Taihu ecosystem has experienced a notable regime shift in the past 14 years. First, nitrogenous nutrients reacted quickly to external emission reductions, showing a notable monotonic decline; in comparison, Total Phosphorus and Cyanobacterial Density followed a non-linear “U-shaped” path, with a notable shift happening in 2020, which marks the change from a “deterioration phase” to a “recovery phase.” Second, correlation analysis has confirmed that the lake is mainly phosphorus-limited, and a clear “decoupling” between nitrogen levels and algal outbreaks has taken place. Third, the “10-year Fishing Ban” (initiated in 2020), together with sustained phosphorus control, reduced the competitive exclusion of phytoplankton by cyanobacteria, promoting the recent rebound in biodiversity. This study points out that Lake Taihu has passed a tipping point of ecological restoration, shifting from a turbid “algae-dominated state” to a stable state with higher biodiversity. Future management strategies should put first the mitigation of internal phosphorus loading and adaptive management against extreme climatic events. Full article

Background/Objectives: To characterize sectoral corneal thickness (CT) profiles in eyes with primary open-angle glaucoma (POAG) compared with healthy eyes and to evaluate potential associations between CT, retinal nerve fiber layer (RNFL) thickness, and optic disc rim area (RA). Methods: In this cross-sectional study, 192 participants (91 with POAG and 101 controls) contributed 297 eyes (145 glaucoma eyes and 152 control eyes). All participants underwent comprehensive ophthalmological examination and spectral-domain optical coherence tomography (OCT; Optovue Solix, Fremont, CA, USA) to obtain peripapillary RNFL measurements, optic disc rim area, and corneal pachymetry maps across five sectors (central, superior, inferior, temporal, and nasal). Repeated-measures correlation analyses were used to assess within-subject associations between CT and RA, and generalized estimating equation (GEE) models were applied to evaluate independent associations between CT, glaucoma status, disease severity, and RNFL thickness while adjusting for relevant covariates. Results: Eyes with POAG exhibited significantly thinner corneas across all sectors compared with controls (all p < 0.05), with the greatest differences observed in the superior (median 607.0 μm vs. 640.0 μm, p < 0.001) and temporal (562.0 μm vs. 579.5 μm, p < 0.001) regions. Average RNFL thickness and rim area were also significantly reduced in glaucoma eyes (all p < 0.001). However, no independent associations between sectoral CT and RNFL thickness or RA were observed after adjustment for multiple comparisons. Although nominal associations between thinner inferotemporal CT and reduced RNFL thickness were observed in unadjusted analyses, these did not remain statistically significant after false discovery rate correction. In multivariable GEE models, glaucoma diagnosis and greater disease severity were consistently associated with reduced RNFL thickness (β range: −11.0 to −42.2 μm; all p < 0.001), whereas CT was not independently associated with RNFL thickness (all adjusted p > 0.07). Conclusions: Sectoral corneal thickness is significantly reduced in eyes with POAG but does not independently correlate with RNFL thickness or optic disc rim area after adjustment for confounding factors. These findings support the concept that corneal thinning reflects structural and biomechanical susceptibility to glaucoma rather than serving as a marker of established neuroretinal damage severity. Further longitudinal studies incorporating comprehensive biomechanical assessments are warranted to clarify the role of corneal structure in glaucoma pathophysiology. Full article

Numerosity can be represented in symbolic formats and non-symbolic dot arrays. How numerosity load unfolds across WM encoding/maintenance and test-stage comparison within a single paradigm remains unclear, especially within the tested 4–6 range. We used a delayed match-to-sample task manipulating numerosity (4–6) and match status, with two test blocks (dot–digit and dot–dot). Behaviorally, a higher numerosity reduced accuracy and increased RTs in both blocks, with larger costs in dot–dot; the mismatch reliably slowed RTs. At sample onset, occipital P1 and N1 amplitudes decreased with increasing numerosity, consistent with greater perceptual/processing demands at higher load, with the strongest differences at the high end of the range. During the delay, numerosity modulation was temporally specific, emerging in the 450–650 ms posterior window and remaining significant after FDR correction across the four consecutive delay windows. At the test, the mismatch elicited a more negative N2 in both blocks (larger in dot–dot), while numerosity also modulated N2 only in dot–dot, showing a monotonic increase in negativity with load. Controlling for condition-mean logRT did not eliminate these N2 effects. P3 showed no reliable modulation, whereas a later positive component was enhanced by mismatch selectively in dot–dot. Together, these results indicate stage-differentiated effects: numerosity load impacts early encoding and a circumscribed maintenance interval, whereas mismatch effects arise primarily during the test-stage comparison, with additional late evaluative activity when formats are aligned. Full article

This study presents a systematic comparative benchmark between two distinct paradigms for solving nonlinear partial differential equations (PDEs): the data-driven Physics-Informed Neural Networks (PINNs) and the analytical Homotopy Analysis Method (HAM). We apply both methods to a unified family of canonical PDEs, the Burgers, Huxley, Fisher, Burgers–Huxley, and Burgers–Fisher equations, under identical problem setups, domain discretization, and validation metrics. PINNs incorporate physical laws directly into neural network training by minimizing a loss function that enforces PDE residuals, yielding physically consistent solutions even for strongly nonlinear problems. HAM provides approximate analytical solutions using a unified framework, and the same initial guess, auxiliary linear operator, and auxiliary function across all equations despite their distinct nonlinearities. The controlled, consistent application of both methods enables a fair, reproducible comparison across this equation family. The results provide a quantitative performance map under identical conditions, delineating when PINNs (high accuracy, long-term stability, and generalization capability) are preferable, versus when HAM (computational speed, short-term analytic approximation, and lower memory footprint) offers advantages. While the finite radius of convergence of the truncated HAM series is theoretically expected, our controlled comparison quantifies for the first time how this degradation varies across equation types, revealing that the choice between methods depends on specific problem requirements including error tolerance, available computational resources, and temporal horizon. The novelty lies not in solving each equation individually, but in deriving a performance taxonomy that systematically connects equation features (shocks, stiffness, and reaction–diffusion coupling) to optimal solver choice—providing previously unavailable, evidence-based guidance for the scientific computing community. This study establishes the first rigorous, controlled comparative benchmark between analytic and data-driven PDE solvers across a spectrum of nonlinearities, providing a reproducible baseline for future hybrid scientific machine learning solvers. Full article