Search Results (1,945)

Background/Objectives: Severe brain injury (SBI) in very preterm infants includes heterogeneous lesions with distinct timing, burden and outcomes. We used cranial ultrasound (CUS) to describe SBI entity, documented timing, three-domain burden, deaths following documented withdrawal, withholding or non-escalation of life-sustaining treatment for poor neurological prognosis (neuro-WWLST), and survivor outcomes. Methods: Retrospective single-center cohort (1991–2020) of 2841 very preterm infants (<32 weeks’ gestation and/or birth weight ≤ 1500 g) with complete CUS within 48 h after birth. CUS was summarized by four windows, three domains—parenchymal lesion, intraventricular hemorrhage (IVH) and ventriculomegaly—and three mutually exclusive entities: periventricular hemorrhagic infarction (PVHI), cystic periventricular leukomalacia (cPVL and grade 3 IVH without PVHI/cPVL (IVH3 entity). Cross-outcome analyses used common maximal-burden CUS. Results: SBI occurred in 286/2841 infants (10.1%) and neuro-WWLST death in 45/2841 infants (1.6%); 43/45 occurred within SBI, and 43/89 SBI deaths (48.3%) followed documented neuro-WWLST. Using common maximal-burden CUS, severe three-domain involvement was more frequent among neuro-WWLST deaths than survivors (37.2% vs. 8.6%). Among SBI survivors with follow-up, cerebral palsy (CP) occurred in 87/176 (49.4%) and clinically classified school-age cognitive sequelae in 50/155 (32.3%). Outcomes varied by entity, with mainly ambulatory unilateral CP after PVHI, more frequent non-ambulatory bilateral CP after cPVL, and a heterogeneous IVH3 profile. Severe three-domain involvement identified a small subgroup with higher outcome burden, but outcomes were not deterministic. Conclusions: A structured, descriptive CUS approach separating lesion entity, documented timing and multidomain burden may support transparent cohort-level description of SBI trajectories, documented neuro-WWLST deaths and survivor outcomes. Full article

The growing complexity of web attacks highlights the need for adaptive, intelligent defense systems that overcome the limitations of traditional rule-based web security. Thus, the architecture proposed in this paper integrates data-driven deep learning with deterministic rule-based logic to enhance real-time detection accuracy and adaptability in dynamic web threat environments. The practical integration of a deep learning-based Gated Recurrent Unit (GRU) model with ModSecurity, an open-source Web Application Firewall (WAF), is employed to improve the detection and classification of malicious HTTP requests. The model, pre-trained on a large labeled up-to-date dataset of web traffic and attack types collected post-2020, is designed to classify requests in real-time, identifying both whether a request is malicious and the corresponding attack category (e.g., SQL Injection, Cross-Site Scripting, Command Injection). We demonstrate how the trained model is incorporated into ModSecurity’s inspection pipeline, allowing it to analyze real-time web traffic alongside traditional rule-based inspection. This hybrid approach aims to significantly reduce false positives and improve adaptability to new attack patterns. Evaluation metrics such as accuracy, receiver operating characteristic (ROC), area under the curve (AUC), Principal Component Analysis (PCA), confusion matrix, and t-Distributed Stochastic Neighbor Embedding (t-SNE) visualization are discussed, along with performance considerations and implementation architecture. The integration presents a robust framework for ML-improved intelligent web security defense. Full article

Traditional global localization systems frequently struggle with perceptual ambiguities in dynamic environments and structurally similar scenes, which severely compromises navigation robustness. Concurrently, conventional path planning methodologies rarely integrate proactive safety considerations regarding high-risk environmental features. To resolve these critical limitations, this paper introduces a comprehensive semantic-assisted framework for mobile robots to enhance both global localization and navigation. First, we develop a semantic-aware place representation derived from LiDAR point clouds. By explicitly filtering dynamic objects and assigning category-specific weights, this approach mitigates perceptual aliasing and ensures robust scene recognition. Furthermore, we implement a Hyper-Semantic Point Histogram (HyperSPH) to embed semantic encoding directly into local geometric features. A Semantic Geometric Consistency Filter is subsequently applied to eliminate matching outliers and maximize registration accuracy. For secure navigation, we propose the Semantic-guided Twin Delayed Deep Deterministic Policy Gradient with Long Short-Term Memory (S-TD3-LSTM) algorithm within a deep reinforcement learning architecture. This strategy extracts temporal correlations via Long Short-Term Memory networks and integrates a dedicated semantic cost function to optimize obstacle avoidance policies. Extensive experiments demonstrate that the proposed localization module achieves superior retrieval and pose estimation precision over conventional methods. In complex path planning scenarios, the S-TD3-LSTM algorithm ensures stable convergence and generates highly adaptive trajectories. By proactively identifying and bypassing semantic hazards, the integrated system drastically minimizes exposure to dangerous zones, successfully establishing a rigorous balance between path efficiency and execution safety. Full article

Stereo vision is essential for passive three-dimensional perception in resource-constrained applications that require low power consumption, predictable latency, and explainable geometry. Although deep learning architectures dominate recent benchmarks, the classical block-matching pipeline remains a foundational approach. Optimizing this pipeline involves navigating complex trade-offs among matching robustness, map density, and computational efficiency. This study systematically surveys and physically validates the classical stereo framework. After revisiting geometric first principles, three matching costs (SAD, NCC, ZNCC) are benchmarked alongside Sobel preprocessing and structural refinements, with subsequent validation using a calibrated consumer webcam rig. Middlebury benchmarks (2001–2021) indicate that while SAD fails under complex radiometric distortion, NCC consistently achieves superior quantitative metrics, incurring only a 1.2-fold computational overhead. Extending the disparity search range improves foreground localization, while block size imposes a trade-off between resolving the aperture problem and preserving fine geometric detail. To bridge theoretical analysis and practical deployment, the pipeline is validated using a custom-calibrated consumer stereo rig. The optimized Sobel-NCC architecture is then evaluated for real-time edge deployment on constrained hardware (NVIDIA Jetson Nano) and narrow-baseline sensors (OAK-D SR) in the context of agricultural robotic manipulation. By prioritizing metric precision over dense prediction, the classical pipeline reconstructs target surfaces with approximately 1 cm depth accuracy at 21 frames per second. These results demonstrate that optimized local algorithms offer deterministic and reliable geometric foundations for real-time edge-computed robotics. Although neural networks are essential for dense reconstructions in ill-posed regions, the foundational principles established here remain indispensable for advanced stereo vision system deployment. Full article

This paper introduces a random activation framework for cure rate modeling that provides a novel latent mechanistic interpretation of the standard mixture cure model, utilizing a Waring-distributed number of latent causes. The proposed approach represents unobserved heterogeneity through a discrete latent variable interpreted as the number of potential risk factors, providing a flexible and biologically interpretable characterization of individual susceptibility. In contrast to classical competing risks models based on extremal operators or deterministic activation schemes, the event time is assumed to arise from a stochastic selection among latent causes. This random activation mechanism defines a unified probabilistic framework in which the cure fraction emerges naturally as the probability of having zero latent causes. The Waring distribution is adopted to model the latent count structure due to its hierarchical formulation, which accommodates overdispersion and heavy-tailed behavior strictly within the latent parametrization of individual risk factors. Under this framework, while the population survival function mathematically reduces to the classical mixture cure representation, the model provides an alternative structure where covariates directly impact the expected latent burden. Parameter estimation for the identifiable regression structure is performed via maximum likelihood, and the finite-sample performance of the estimators is assessed through Monte Carlo simulations, showing accurate parameter recovery and stable inferential properties. An application to real survival data illustrates the practical relevance and epidemiological interpretability of the proposed framework. Overall, this work extends the understanding of existing cure rate models by integrating latent count structures and stochastic activation within a coherent setting, providing a powerful interpretation tool for heterogeneous survival data with long-term survivors. Full article

Integration of LiDAR and thermal sensing has become increasingly important in robotics, infrastructure diagnostics, environmental monitoring, and autonomous perception systems. LiDAR sensors provide accurate three-dimensional geometric information but do not directly capture thermal properties of observed objects, whereas thermal cameras provide temperature distributions without explicit spatial structure. Fusion of both sensing modalities enables thermally augmented 3D scene reconstruction and spatial localization of temperature anomalies. This paper presents a practical LiDAR–thermal fusion framework for three-dimensional localization of heat sources using an Ouster OS1 LiDAR sensor and a FLIR A70 thermal camera. The proposed framework includes intrinsic thermal-camera calibration, extrinsic LiDAR–thermal calibration, multimodal data synchronization, projection of LiDAR points onto the thermal image plane, and assignment of temperature values to spatial points. Additionally, a dedicated thermally distinguishable calibration target is proposed to enable reliable multimodal feature extraction under low-contrast LWIR imaging conditions. The developed framework was experimentally validated using real radiometric thermal data and LiDAR point clouds acquired under laboratory conditions. Quantitative evaluation demonstrated reprojection errors below 1 pixel and a mean hottest-point localisation error of approximately 4.1 cm at a distance of 12.3 m. The results confirm that accurate spatial localisation of thermal anomalies can be achieved using a geometry-based multimodal fusion approach without relying on computationally expensive learning-based methods. The proposed framework emphasises practical deployment, deterministic calibration, and applicability in scenarios where limited training data or constrained computational resources make learning-based approaches difficult to apply. The proposed system may be applied to building energy diagnostics, industrial inspection, technical infrastructure monitoring, and robotic perception systems that require reliable spatial localisation of heat sources under real measurement conditions. Full article

This paper presents a two-stage stochastic unit commitment model to mitigate the costs and reliability effects of high wind energy penetration into the power system. Wind energy variability/uncertainty is explored through this system. Mixed-integer linear programming (MILP) is used to solve the model and find optimal unit commitment and dispatch variables under uncertainty in wind conditions. The model champions reduced reliance on deterministic approaches, lowered costs, increased wind utilization, and provides reliable systems that can sustain 24 h. Wind energy is expected to constitute a significant share of future electricity generation portfolios; however, its inherent intermittency and variability often lead to mismatches between energy supply and demand. This uncertainty complicates generation scheduling decisions, particularly in determining which power plants to commit, their operating durations, and the optimal dispatch timing. Consequently, advanced optimization strategies are required to ensure efficient and reliable system operation. The proposed approach provides a structured, robust framework for optimal generation scheduling and resource allocation, even under limited wind availability. By enhancing the integration of wind energy into the power system, the method minimizes operational costs, improves resource utilization, and maintains system reliability and stability. Full article

This paper investigates machine learning (ML)-supplemented workflows integrated within a modular parametric modelling framework derived from a typological analysis of stadiums. The objective of the research is to address a gap between numerous isolated computational studies and the realities of early stadium design within the Royal Institute of British Architects (RIBA) Plan of Work (PoW) Stages 0–3. From a practical perspective, the proposed design framework aims to embed supervised learning, semi-supervised learning, and evolutionary optimisation into stadium design development to support site appraisal, brief preparation, concept development, spatial coordination, and stadium bay or stand optimisation based on quantifiable design characteristics. The framework addresses the inefficiencies and limitations of the traditional stadium design process by allowing rapid design space exploration defined by typological drivers, evaluation of a large set of solutions based on performance metrics such as circulation distances, sightline quality, and layout distribution, and the validation of concepts against benchmarks. Within the applicable design pipelines, and where labels are derived from deterministic performance criteria, the supervised approaches achieved prediction accuracies above 85%, while evolutionary optimisation reduced the number of seats with restricted views by approximately 95%. The value of the study is that it demonstrates that the integration of parametric modelling based on shared typological characteristics and the mapping of ML methods to the RIBA PoW has the potential to support stadium design in a novel way. Full article

Accurate channel state information is essential for the performance of modern wireless communication systems. Conventional channel estimation typically relies on uplink Sounding Reference Signals (SRSs), which can introduce considerable overhead and power consumption, particularly in high-mobility or resource-constrained scenarios. To alleviate this burden, this paper explores an alternative approach that leverages sensing channel information to assist communication channel reconstruction. A shared cluster concept is introduced to capture the correlation between sensing and communication channels, and a sharing probability function is derived through statistical analysis of ray tracing simulation data across multiple scenarios. The shared cluster parameters extracted from the sensing channels are integrated into a cluster-based channel modeling framework to reconstruct the downlink communication channel. A deterministic simulation platform is developed using the Sionna ray tracing library, and the K-Power-Means algorithm is employed for multipath clustering. Simulation results demonstrate that the reconstructed channel closely matches the original channel in terms of the power delay profile and the root mean square delay spread, with mean values of 84.16 ns and 73.52 ns, respectively. The proposed method offers a promising supplementary approach for channel acquisition in scenarios where frequent SRS transmission is undesirable, and provides insights for future sensing-assisted communication system design. Full article

Shannon entropy is the most common measure that one could use to check if a data source has random behaviour or not. A value close to the maximum is usually considered as evidence that the source is “random enough”. The present paper shows that this criterion alone is not enough. A deterministic logistic map driven at $r=3.9999$ reaches 94.97% of the Shannon maximum, yet it is fully predictable once we look at the built-in patterns: its permutation entropy drops to 77.01% of the maximum and its sample entropy falls to 0.67, against 2.33 for a high-quality pseudo-random generator (PRNG). Building on this observation, we combine four entropy measures—Shannon, Rényi, permutation, and sample—into a single diagnostic profile of the analyzed source. In order to validate our approach with practical, real life data, we test it on 2538 official draws of the Romanian Loto 6/49 lottery, recorded between August 1993 and April 2026. The lottery historical data set is very close to a high-quality PRNG (pseudo-random number generator) from the point of view of all four measures. We also observe that the entropy deficit of both the lottery and the PRNG decays as a power law with exponent $\alpha \approx -0.96$; in contrast, the logistic map sits at $\alpha \approx -0.07$. A Random Forest classifier trained only on the entropy profile reaches 78% accuracy on the analyzed four-way classification task (lottery, PRNG, logistic map, biased distribution), but scores 55.7% on the binary lottery-versus-PRNG task, consistent with chance. The method introduced in this study is domain-independent and applies directly to RNG certification, cryptographic key auditing, and any setting where structured pseudo-randomness has to be ruled out. Full article

Exchange-rate forecasting requires modelling non-stationary dynamics, heavy-tailed shocks, and complex temporal dependencies. However, forecasting performance in emerging-market currencies is fundamentally constrained by intrinsic dynamical instability, while most existing approaches are evaluated primarily through predictive accuracy rather than the predictability limits of the underlying system. This paper develops a predictability-aware framework that combines nonlinear dynamical diagnostics with a Lévy-driven neural stochastic differential equation model. Drift and diffusion are parameterized by neural networks and driven by $\alpha$-stable Lévy motion, enabling the representation of non-Gaussian fluctuations, abrupt shocks, and regime changes. To learn under discontinuous dynamics, we introduce a structurally constrained training objective based on a strong-form discretization of the underlying SDE. To characterise intrinsic predictability, we employ phase-space reconstruction and maximal Lyapunov exponent estimation. These diagnostics are interpreted as finite-sample measures of trajectory divergence and effective instability in a stochastic system, rather than evidence of low-dimensional deterministic chaos—a distinction motivated by well-documented limitations of chaos testing in financial data. Experiments on multiple West African currency pairs demonstrate competitive short-horizon forecasting performance relative to econometric and neural baselines while providing a principled framework for analysing predictability degradation under heavy-tailed stochastic dynamics. Across currencies and model classes, forecasting accuracy deteriorates beyond horizons comparable to the estimated Lyapunov time, suggesting that forecast degradation reflects intrinsic dynamical instability rather than model-specific limitations. The results support the view that reliable exchange-rate prediction is fundamentally a short-horizon problem and illustrate how stochastic dynamical modelling and predictability diagnostics can be combined to characterise forecasting limits in heavy-tailed financial systems. Full article

This paper presents a full-scale case study on real-time corrosion monitoring in an underground reinforced-concrete potable water tank built in 1968. The study aims to demonstrate how continuous electrochemical monitoring can support durability assessment and predictive maintenance in ageing water-retaining infrastructure, where direct inspection is often limited and exposure conditions are spatially variable. Fourteen monitoring points were installed in beams, columns and domes subjected to different exposure conditions. Corrosion potential, concrete resistivity, corrosion current density and temperature were recorded every 3 h and used to assess the corrosion state of the reinforcement. The monitored durability indicators were reinforcement section loss, estimated from corrosion current density using Faraday’s law, and corrosion-induced crack-width evolution, used as a serviceability-related indicator for maintenance planning. The results show that beams remained predominantly passive, with corrosion current densities below 0.1 µA/cm² and incremental sectional losses below approximately 2 µm during the monitoring period. Columns showed the highest vulnerability, particularly at lower elevations subjected to prolonged immersion, with estimated incremental section losses reaching approximately 4–6 µm and a clear correlation between submerged time and corrosion progression. Domes exhibited intermediate behaviour, with occasional activation events associated with environmental fluctuations. A multivariable model combining resistivity and temperature was used to interpret corrosion kinetics, while Faraday-based section-loss estimates were coupled with empirical crack-width models to forecast serviceability indicators up to 2045. These forecasts are presented as scenario-based maintenance-support indicators rather than deterministic predictions of future damage, since corrosion propagation and crack development may evolve nonlinearly under changing exposure conditions. The proposed approach demonstrates how continuous corrosion monitoring can be linked to durability limit-state assessment, enabling risk-informed and performance-based maintenance of critical water infrastructure. Full article

Horizontal-well steering fracturing is an important completion strategy for unconventional reservoirs, where fracture growth is jointly controlled by wellbore azimuth, natural fractures, and inter-cluster stress interference. In this study, a two-dimensional fluid-solid-coupled hydraulic-fracturing model with embedded cohesive elements was developed to simulate fracture initiation and growth at steering angles of 0°, 30°, 45°, 60°, and 90°. The Blanton, Warpinski-Teufel, and Blanton-Gao hydraulic-fracture/natural-fracture interaction criteria were used as mechanical benchmarks to interpret simulated capture, deflection, and penetration regimes. The simulations indicate that natural fractures preferentially guide fracture propagation: hydraulic fractures tend to be captured by, or propagate along, natural fractures at approach angles ≤30°, whereas penetration is more likely at approach angles ≥60°. In the single-stage single-cluster model, the 90° case produces the largest simulated fracture length and the highest failed-cohesive-element count. In the single-stage multi-cluster model with 3 m cluster spacing, the 30–45° interval shows more favorable fracture extension and interface activation than the 90° case because inter-cluster stress-shadow effects suppress fracture-network development at large steering angles. The resulting steering-angle window should be interpreted as a comparative result for the fixed mesh, deterministic natural-fracture realization, and baseline cluster-spacing configuration adopted here. These results provide a mechanistic basis for steering-fracturing design in hard-rock reservoirs while clarifying the applicability limits of the two-dimensional plane-strain approximation. Full article

This study presents a hybrid spatio-textual matching approach for integrating historical web-derived address datasets with a municipal reference dataset, using the 2009 administrative delineation of Şişli (Istanbul) as a case study. The proposed approach addresses challenges commonly encountered in data obtained from web archives, such as lexical differences, abbreviations, heterogeneous structures, and missing address information. The methodology consists of three main stages: (i) preprocessing and structuring of web-based address records; (ii) hybrid matching, combining deterministic rules with similarity-based methods; and (iii) post-matching geographic enrichment using an Application Programming Interface (API) to provide supplementary geographic context for matched records. The matching process is conducted exclusively between historical datasets; contemporary geographic information is used only after the completion of the matching process to provide additional contextual information. The methodology integrates token-based, vector-based, and structural similarity measures within a calibrated scoring scheme to improve the matching of ambiguous and inconsistent address records. The results indicate that 65.4% of the records were automatically accepted, 7.3% required manual review, and no suitable candidate was found for 5.4%. Deterministic matching results reveal that strict rule-based approaches are highly sensitive to data integrity and attribute consistency, especially in heterogeneous web-based datasets, highlighting the value of combining multiple similarity measures within a hybrid matching strategy. The API-based enrichment results provide supplementary geographic context regarding the contemporary surroundings of matched records, while historical interpretations remain grounded in the original archival datasets. In this context, the study may contribute to the integration of historical web-based address data with structured municipal datasets under heterogeneous archival data conditions. Full article

Urban Unmanned Aerial Vehicle (UAV) logistics operations are frequently constrained by the intersection of limited battery endurance and dynamic ground traffic. When mobile depots are delayed by congestion, onboard drone fleets experience extended idling periods, leading to constrained sortie generation and reduced asset utilization. To address this bottleneck, this paper introduces a traffic-predictive multi-UAV dispatch framework for deterministic day-ahead planning under modeled urban operating conditions. By coupling a count-derived macroscopic speed surrogate learned using XGBoost with a Particle Swarm Optimization (PSO)–Mixed-Integer Linear Programming (MILP) optimization architecture, the framework synchronizes mobile depot trajectories with forecasted low-congestion windows and pre-allocates endurance-feasible parallel aerial sorties. Controlled computational experiments across 30 synthetic routing instances demonstrate the potential value of this approach within the stated modeling assumptions. Compared to baseline clustered deployments, the traffic-aware framework raises mean fleet utilization from 0.43 to 0.63—a 46.2% relative improvement driven by temporal compression of the mission window rather than an absolute increase in flight hours. Furthermore, the proposed framework reduces total mission completion time by 69.87% relative to the conventional truck-only baseline, while achieving a 29.58% incremental gain over static speed drone deployments. These findings suggest that incorporating predictive ground traffic information into day-ahead UAV scheduling can improve modeled fleet efficiency; however, field validation with measured route-level speeds, real delivery demand, and operational constraints remains necessary before deployment-level claims can be made. Full article