Search Results (500)

Background: Dyslipidemia is becoming an increasingly important issue from a public health point of view and is generally treated with statins. However, as generic versions of these drugs are now widely available, doubts have emerged about their effectiveness. The current study examines the clinical effectiveness and impact on the national budget of using generic atorvastatin compared to the branded version. Methods: A retrospective comparative study was conducted at King Khalid University Hospital between 2015 and 2025, involving 286 patients (142 receiving the generic form and 144 receiving the branded form). Changes in lipid panel values were assessed using a Generalized Linear Model (GLM) to account for multiple factors. The economic outcomes were estimated using probabilistic (10,000 Monte Carlo iterations) and deterministic scenario analyses. Results: The mean adjusted decrease in low-density lipoprotein (LDL) was 26.11% in the generic group and 19.34% in the branded group. After adjusting for multiple variables, the branded group showed a non-significant lower reduction in LDL (β-estimate: −0.255, 95% CI: −0.119 to 0.391, p = 0.3318). The Monte Carlo simulations showed large projected mean annual savings when using the generic formulation: $29,717,778 based on tender pricing and $295,432,159 based on retail pricing. The deterministic sensitivity analyses consistently found positive projected savings across all population and market-share scenarios tested. Conclusions: Generic atorvastatin was found to have similar clinical effectiveness to the branded version in terms of reducing atherogenic lipids, with no statistically significant difference between the two groups. Furthermore, the economic modeling shows that using generic atorvastatin yields considerable national-level cost savings. Full article

Lithium aluminosilicate (LAS) ultra-low-expansion glass ceramics are core materials for precision optical systems, whose quality and dimensional stability are critically constrained by the uniformity of the coefficient of thermal expansion (CTE). This work proposes a nondestructive ultrasonic immersion pulse reflection (UIPR) method for rapid and low-cost characterization of the CTE in LAS glass ceramics. Key parameters of the ultrasonic measurement system are optimized via finite element method (FEM) simulations and experimental validation. Employing the correlation method, the ultrasonic longitudinal wave velocity is measured in LAS glass ceramic samples with distinctly different CTE values. The proposed method achieves an ultrasonic longitudinal wave velocity measurement uncertainty of 0.49 m/s, contributing 4.78 ppb/°C to the overall uncertainty of ultrasonic CTE determination. Within the investigated sample set, a negative relationship is observed between ultrasonic longitudinal wave velocity and the mean CTE (0–50 °C). The linear fit yields a slope of −9.75736 (ppb/°C)/(m/s), with a Pearson correlation coefficient of −0.84303. Featuring noncontact and nondestructive capabilities, this method lays a solid methodological foundation for CTE evaluation and efficient iterative optimization of material fabrication processes. Meanwhile, it shows great potential for rapid full-aperture characterization of CTE uniformity in large-size LAS glass ceramics. Full article

Financial risk along the China–Pakistan Economic Corridor may combine fast firm-level scaling changes with slower joint-tail exposure. This study evaluates a leakage-controlled two-speed architecture for seven Pakistani equities during July 2021–June 2026. Causally timed, bounded-influence wavelet p-leader states feed quantile learners, while 13 prespecified directed hyperedges define a structural map. Evidence comprises 999 iterative amplitude-adjusted Fourier-transform (IAAFT) surrogates per node, 999 matched random edge sets, a 249-origin locked Value-at-Risk–Expected Shortfall test, learner-by-feature ablations, moving-block inference, and cost-adjusted portfolios. Robustification removes the legacy KEL spectrum anomaly; only LUCK rejects the IAAFT null after false-discovery-rate control (q = 0.009). No hyperedge has a confirmed increment beyond singleton conditionals and the pairwise-only benchmark. The map’s mean lift is 0.951 versus a placebo median of 0.989 (p = 0.834). The reservoir is competitive in point loss but indistinguishable from the conditional autoregressive Value-at-Risk benchmark (p = 0.420); its p-leader gain does not transfer across learners or survive KEL-block exclusion. The 0.606-percentage-point drawdown difference relative to historical conditional Value-at-Risk (CVaR) is imprecisely estimated (95% interval: [−7.652, 5.648]; p = 0.626). The framework supports robust multiscale measurement and transparent structural mapping, not general forecasting superiority, unique hyperedge information, or reliable portfolio protection. Full article

Survival analysis plays a central role in medical research. Although the Cox proportional hazards (CoxPH) model remains the standard approach, machine learning and deep learning methods have been increasingly adopted. However, many published comparisons have relied on a single train–test split, which may produce unreliable performance estimates, particularly for unstable modelling approaches. This study compared CoxPH (LASSO-selected), Random Survival Forest (RSF), and Long Short-Term Memory (LSTM) networks using four survival datasets: breast cancer ($n=4024$), heart failure ($n=299$), recidivism ($n=4618$), and the Mayo Clinic Primary Biliary Cholangitis Sequential Dataset (PBC2; $n=312$) containing time-varying covariates. Model performance was evaluated using a two-stage protocol comprising a conventional 80–20 stratified train–test split followed by 100 repeated stratified 80–20 train–test splits. Performance was assessed using the concordance index (C-index), integrated Brier score (IBS), and time-dependent area under the receiver operating characteristic curve (AUC), with statistical significance determined through distributional assumption testing, adaptive omnibus tests, and Bonferroni-adjusted pairwise comparisons. For the three static datasets, single train–test splits suggested moderate LSTM performance (C-index: 0.65–0.70); however, repeated evaluation showed that this finding was not robust. Across 100 iterations, CoxPH and RSF consistently outperformed LSTM (all $p<0.001$), achieving mean C-index values ranging from 0.66 to 0.73 compared with 0.31 to 0.42 for LSTM, with very large effect sizes (Cohen’s d: 6–30). In contrast, on the longitudinal PBC2 dataset, the LSTM-based model achieved the highest repeated C-index ($0.806\pm 0.040$), compared with $0.777\pm 0.042$ for CoxPH and $0.779\pm 0.040$ for RSF, with only 1.2% performance degradation under repeated evaluation. These findings indicate that reliance on a single train–test split can produce unstable and potentially unrepresentative estimates of model performance. Traditional survival models were more accurate and stable for datasets containing only static baseline covariates, whereas the LSTM-based model showed higher discrimination on longitudinal survival data with genuine temporal structure, though this advantage is confounded with greater access to patient history and cannot be attributed to architecture alone. Overall, the results underscored the importance of repeated evaluation as a more reliable framework for comparing survival models, and are consistent with, though do not conclusively establish, the importance of matching model architecture to data characteristics. Full article

The Berca–Arbănași region of Buzău County (Romania) is of exceptional geochemical interest due to diapiric structures, active mud volcanoes, and historical subsurface hydrocarbon deposits. This study presents the first comprehensive geochemical assessment of surface soils from the Berca–Arbănași Subcarpathian zone, based on the analysis of 27 soil samples collected along three north–south transects and the determination of 12 potentially toxic elements (As, Cd, Co, Cr, Cu, Fe, Hg, Mn, Ni, Pb, V, Zn) using aqua regia digestion and ICP-MS. The local geochemical background was calculated using the iterative median ± 2MAD method. Twelve pollution and ecological risk indices were computed: the Pollution Index (PI), Contamination Factor (CF), Geoaccumulation Index (Igeo), Enrichment Factor (EF), Pollution Load Index (PLI), Modified Degree of Contamination (mCd), Nemerow Integrated Pollution Index (PINemerow), Ecological Risk Factor (E_rⁱ), Ecological Risk Index (RI), Mean Effect Range-Median Quotient (MERMQ), Degree of contamination (Cdeg), and the V/Ni petroleum origin indicator. Results show that 66.7% of samples are classified as polluted (PLI ≥ 1; mean = 1.102), with moderate enrichment in Cd, Cu, Hg, Pb, and Zn, attributable to diffuse anthropogenic sources. The ecological risk index (RI) remains low across all samples (mean = 35.96; all < 150), indicating that, despite moderate pollution, ecological risk is currently low. The V/Ni ratio (0.391–0.884, mean = 0.608) is below 1.0 for all samples, indicating a lithogenic (not petroleum) origin of vanadium and nickel and confirming a negligible geochemical impact of mud volcanoes and hydrocarbon extraction activities in the area at the sampled locations. The study establishes baseline geochemical reference values for the Berca–Arbănași area and provides data for sustainable land management, environmental monitoring, and evidence-based policymaking. These results directly support the objectives of the EU Soil Strategy 2030 and align with the United Nations Sustainable Development Goals on food security (SDG 2), good health and well-being (SDG 3), and life on land (SDG 15). Full article

Background: Endometriosis affects approximately 6–10% of women of reproductive age—an estimate that varies substantially with the diagnostic standard applied—and is increasingly recognized as a systemic inflammatory disease. Growing evidence implicates shared immunological pathways between endometriosis and autoimmune thyroid disease, yet large-scale real-world evidence from standardized multi-site data remains limited. We evaluated the association between endometriosis and newly recorded autoimmune thyroid disease using federated Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) data from 12 Korean hospitals. Methods: We conducted a propensity score (PS)-matched cohort study using OMOP-CDM version 5.3 data from 12 Korean tertiary academic medical centers. Women aged 18–60 years with a first recorded endometriosis diagnosis were matched 1:1 to controls without endometriosis on age, calendar year, hypertension, and selected Charlson comorbidity index components. Site-specific Cox proportional hazards models, fitted to patient-level records locally at each institution, estimated hazard ratios (HRs) for newly recorded autoimmune thyroid disease, defined as Hashimoto’s thyroiditis or Graves’ disease. Pooled estimates were derived using DerSimonian–Laird random-effects meta-analysis; leave-one-out meta-analysis, meta-regression on site-specific follow-up ratio, and an E-value were used to assess robustness. Results: After 1:1 PS matching, 71,619 women with endometriosis and 71,619 matched controls were included. Mean follow-up was 2693 days in the endometriosis group and 1677 days in the control group. A directionally positive but statistically inconclusive association was observed between endometriosis and autoimmune thyroid disease (pooled HR 1.22, 95% CI 0.97–1.53; p = 0.090; I² = 38.2%). Seven of twelve sites reported HRs above 1.0, including two sites reaching individual statistical significance: AUMC (HR 1.38, p = 0.030) and KHUH (HR 2.69, p < 0.001). The pooled HR remained above 1.0 in all 12 leave-one-out iterations (range 1.17–1.31); omission of KHUH reduced I² to 6.6%. Site-specific follow-up imbalance was not associated with the site-specific log HR (meta-regression p = 0.887). The E-value for the point estimate was 1.74. Conclusions: In this multi-site OMOP-CDM analysis, endometriosis showed a directionally positive but statistically inconclusive association with newly recorded autoimmune thyroid disease. The confidence interval is compatible with both no association and a clinically meaningful increase in risk, and the findings are hypothesis-generating rather than confirmatory. They do not support any change in the clinical evaluation of women with endometriosis. Further validation is warranted using standardized outcome definitions, thyroid autoantibody measurements, and thyroid function tests. Full article

Accurate prediction of automotive aerodynamic wind noise is important for cabin comfort and early-stage styling, yet conventional CFD and wind-tunnel workflows are too expensive for rapid design iteration. This paper proposes a point cloud surrogate that combines farthest-point sampling with a Transolver-derived, physics-inspired slice-attention mechanism. Here, physics-inspired denotes a representation-level inductive bias; the model does not impose governing-equation residuals, conservation constraints, or physics-based losses. Exterior meshes are converted into 10,240-point geometric inputs and assembled into a controlled dataset of 867 sedan and SUV variants generated at 120 km/h and zero yaw. On the random test split, the model obtains RMSE values of 2.30 dB(A), 2.56 dB for SPL, and 0.0068 for the dimensionless articulation index (AI), with 0.80 s single-sample inference on an RTX 4090. Repeated-seed and grouped-split analyses indicate a favorable accuracy–latency trade-off while also showing a measurable performance decrease for held-out vehicle families. A single-vehicle wind-tunnel comparison confirms strong frequency-trend correlation but reveals a mean simulation over-prediction of 2.70 dB; therefore, the current surrogate should be interpreted primarily as an emulator of the simulation labels rather than a universally unbiased predictor of measured cabin noise. Full article

To address the performance degradation of target elevation estimation for deep-sea vector hydrophone vertical arrays under strong interference, this paper proposes an inverse beamforming (IBF)-based interference suppression method for such vertical arrays. The proposed method exploits the array spatial response characteristics to detect, reconstruct, and iteratively cancel dominant interference components. An adaptive stopping strategy is adopted to prevent excessive cancellation and reduce the risk of target self-cancellation. Furthermore, it utilizes the coherence properties between multiple physical channels to perform joint processing of sound pressure and particle velocity channels, thereby enhancing the target-direction response and mitigating residual interference. To alleviate the performance degradation induced by array element failures, this paper introduces a least-squares-based array output reconstruction method to restore the spatial sampling structure and reduce array-manifold distortion. Simulation and sea trial results demonstrate that, compared with conventional beamforming (CBF) and minimum variance distortionless response (MVDR), the proposed method effectively suppresses strong directional interference and recovers the target-direction spatial response. The median spatial-spectrum contrast between the target region and the interference region improves from −1.21 dB before IBF to 2.70 dB after IBF, with a median improvement of 4.12 dB over the full observation interval. The mean estimated elevation angle is 46.5°, close to the reference value of 47.2°, with an RMSE of 0.76° and a success rate of 100% within ±2° and ±3° tolerances. These results indicate that the proposed method enables reliable target elevation estimation under strong interference and remains effective in the presence of array element failures, demonstrating its suitability for practical underwater applications. Full article

Optical systems in the real world often create image problems, such as Gaussian blur from defocus or atmospheric turbulence, and radial vignetting caused by lens shape. Standard mixed-data fine-tuning helps the task head handle these issues, but it does not actually fix them. We introduce the Iterative Correction of Optical Perturbations ICOP framework, which corrects encoder feature representations before they reach the task head using a physics-aware plug-in module. ICOP does this by modeling blur as an isotropic-Gaussian point-spread function (PSF) and uses gradient-based, self-supervised optimization (Adam) to discover feature-space corrections that steer degraded representations toward their clean-data distribution. It includes a BlurEstimator that builds a degradation descriptor using fixed Laplacian and Sobel operators, and a PromptGenerator that turns this descriptor into modulation parameters for the frozen encoder output. The framework comes in two versions based on the task: an additive correction (ICOP-Add) for image classification, and a Feature-wise Linear Modulation correction (ICOP-FiLM) for object detection. We observe a convergence between clean-task performance and blur-induced degradation across datasets, consistent with greater reliance on high-frequency features in stronger backbones; we treat this as an empirical, cross-dataset observation rather than a demonstrated causal claim (task difficulty, category structure, texture, and object scale also differ across datasets). Independently of this, ICOP-FiLM’s corrective benefit does not scale with degradation severity, revealing a more nuanced relationship between backbone quality and robustness. For classification, ICOP-Add improves distorted-condition accuracy over strong mixed fine-tuning by $+2.4$, $+9.1$, and $+10.7$ percentage points on MNIST, FashionMNIST, and CIFAR-10, respectively (McNemar’s test, $p<0.001$ on all three, 5000 paired predictions per dataset). On three object detection datasets, ICOP-FiLM improves distorted-condition mAP over a mixed-fine-tuning null hypothesis by $+0.026$, $-0.005$, and $+0.003$ mAP, respectively (all values mean over 3 seeds). Against a matched-blur-ratio control that isolates the correction module’s own contribution, ICOP-FiLM wins by a consistent margin on two of the three datasets ($+0.037$ and $+0.024$ mAP, winning in every one of 3/3 seeds on each) and loses on the third ($-0.039$ mAP, losing in 3/3 seeds); it outperforms parameter-efficient (VPT, Adapter) baselines trained on identical data on the same two datasets. This dataset-dependent pattern is discussed in detail in the main text. ICOP-FiLM adds only 82,672 parameters to a 42-million-parameter Real-Time DEtection TRansformer (RT-DETR) detector. All reported results are obtained under synthetic Gaussian blur and radial vignetting applied to clean images from the six benchmark datasets studied. Full article

LiDAR–inertial odometry (LIO) is accurate in structurally rich environments but can become weakly observable in tunnels, stairways, and open terrain. This study introduces a degeneracy-aware, intensity-assisted LIO method that uses LiDAR reflectivity as an internal sensing modality without requiring a camera. Raw returns are projected onto a normalized panoramic intensity image, and image patches are selected according to their ability to complement the uninformative directions identified from the geometric information matrix. Point-to-plane, photometric, and inertial residuals are then fused in an iterated extended Kalman filter. Unlike fixed-scale intensity fusion, the proposed strategy adjusts the photometric residual scale according to the number of weak geometric directions and smooths this scale temporally. Experiments on the Newer College, ENWIDE, and GEODE datasets show that adaptive scaling reduces translational ATE RMSE, defined as the sequence-level root-mean-square of pose-wise translational absolute pose errors, by 11–40% on four ablation sequences. On GEODE-Stairs, the method obtains an ATE RMSE of 0.26 m, which is 46.9% lower than that of COIN-LIO. It also achieves the lowest ATE RMSE on each of the four tunnel sequences, with values ranging from 0.30 to 0.33 m. Mean processing times range from 15 to 25 ms per scan, corresponding to an average throughput of 40.0–66.7 scans/s on the evaluated platform. These results indicate that degeneracy-conditioned intensity fusion improves LIO robustness while maintaining average throughput compatible with real-time operation. Full article

Metaheuristics require sustained global search without sacrificing local refinement, yet many variable-structure methods change operators through one-way iteration schedules. We introduce the Weather State Ants Optimizer (WSAO), in which a discrete-time Markov chain recurrently selects one of three population updates. Sunny, cloudy, and rainy states correspond to global exploration, movement toward nests, and local refinement, respectively. An archive-based mechanism also maintains several spatially separated nests as concurrent search centers. Thirty independent runs compared WSAO with 11 algorithms on 29 CEC2017 and 12 CEC2022 functions. WSAO achieved the lowest Friedman mean rank on both suites, at 2.48 and 2.33. Across five constrained design cases, it joined the leading group by mean objective value on four cases and ranked second on pressure-vessel design. Targeted CEC2022 controls showed that no alternative transition matrix dominated the baseline. Eliminating the trial perturbation worsened every selected function, whereas the contribution of multiple nests depended on the landscape structure. The combined evidence supports recurrent state-controlled search as a competitive framework for continuous numerical and constrained optimization. Full article

Sustainability Education (SE) is a complex pedagogical challenge that involves multiple disciplines and requires consensus on the content to be prioritized for citizenship education. Consensus-building on such complex issues can be effectively achieved through the Delphi technique, which relies on an anonymous panel of experts providing assessments through an iterative process. This study employed a modified Delphi method to develop a core curriculum for SE, tailored for non-specialist students at the pre-university or undergraduate level. The process involved two stages: first, semi-structured interviews were conducted with 10 experts to generate a list of key topics; subsequently, a questionnaire was administered using a ten-point Likert scale covering 59 key concepts, organized by conceptual area. The second panel, consisting of 40 professors from 10 departments at the University of Trento, evaluated and selected the topics for the curriculum. A consensus value was calculated for each item, and concepts were ranked according to their perceived importance and level of agreement. This research aims to arrive at a collective assessment that goes beyond individual disciplinary perspectives. The results demonstrate a strong multidisciplinary consensus on foundational and action-oriented dimensions, with the highest priority assigned to environmental sustainability and systemic interdependence (mean score of 9.3 on a 10-point scale), social equity (9.1), and operational energy efficiency (8.8). Conversely, highly specialized tools, such as carbon markets (7.2) and climate psychology (7.0), were deemed less central for general education. Furthermore, the analysis reveals that experts’ disciplinary backgrounds profoundly influenced their priorities, underscoring the necessity of a truly transdisciplinary approach. The results provide a systematic foundation for the development of interdisciplinary educational programs. Full article

In electric vehicles (EVs), the battery management system (BMS) plays a central role in ensuring safe, efficient, and reliable battery operation under varying driving and environmental conditions. The effectiveness of a BMS largely depends on the availability of an accurate battery model, whose performance is strongly influenced by the precision of its identified parameters. However, estimating these parameters remains a difficult nonlinear optimization problem, especially under low state of charge (SOC) operation. Classical identification approaches often have limited robustness under such conditions, while metaheuristic algorithms provide a promising alternative because of their ability to handle nonlinear and multimodal search spaces. Even so, many existing methods still encounter drawbacks related to convergence speed and susceptibility to local optima. Motivated by these challenges, this study investigates the recently introduced Artificial Lemming Algorithm (ALA) for parameter identification of a second-order equivalent circuit model (2RC-ECM) under EV-oriented low-SOC operating conditions. Experimental validation is conducted using two independent dynamic datasets, namely the High Dynamic Profile (HDP) at 25 °C and the Urban Dynamometer Driving Schedule (UDDS) at −5 °C, involving different lithium-ion cells and operating conditions. ALA is benchmarked against nine competing metaheuristic algorithms under identical search boundaries and computational settings. Performance is assessed using RMSE, MAE, MaxAE, bias, convergence behavior, error distributions, execution time, and sensitivity to the number of independent runs, population size, and maximum number of iterations. The results show that ALA achieves the lowest minimum, mean, and maximum RMSE for both datasets, with minimum RMSE values of $0.01075 \mathrm{V}$ for HDP and $0.03534 \mathrm{V}$ for UDDS. Unseen-data validation further yields RMSE and MAE values of $0.0082$ and $0.0061 \mathrm{V}$, respectively, for HDP, and $0.0416$ and $0.0299 \mathrm{V}$, respectively, for UDDS. In addition, convergence, error-distribution, and sensitivity analyses show that ALA maintains competitive and consistent performance across the investigated configurations. Overall, the results demonstrate that ALA provides a favorable balance between estimation accuracy, robustness, convergence behavior, and computational cost for offline lithium-ion battery parameter identification. Full article

The Coordinate Rotation Digital Computer (CORDIC) algorithm is widely employed in digital signal processing and hardware accelerators because it computes a broad range of elementary functions using iterative shift-and-add operations. Conventional CORDIC implementations, however, execute a fixed number of iterations irrespective of the input characteristics or the precision required, resulting in unnecessary computational overhead and increased execution latency. This work presents an explainable machine learning framework for adaptive iteration optimization in a multi-mode CORDIC architecture supporting circular, hyperbolic, and linear operating modes. A unified prediction framework for calculating the optimal number of iterations is made possible by the developing a generic feature representation to describe the numerical behavior of CORDIC computations across various modes. We systematically evaluated eight regression models, including Linear Regression, Decision Tree, Random Forest, Extra Trees, Support Vector Regression, Multi-Layer Perceptron, and Extreme Gradient Boosting (XGBoost) and LightGBM. Among the models evaluated, the Decision Tree achieved the best performance on an independent test set of 2305 samples from 461 previously unseen input groups, with a MAE of 0.9160 iterations, RMSE of 1.9671, and $R^2$ of 0.6076. Predictions were within one and two iterations of the reference value for 80.26% and 90.07% of the test samples, respectively. Since prediction accuracy alone does not guarantee that the required numerical tolerance will be satisfied, the predicted iteration count was further evaluated using the actual CORDIC error, followed by a safety-correction procedure. The safety-corrected approach achieved 100% tolerance satisfaction on the independent test set, reducing the mean number of iterations from 20 to 11.739, corresponding to a 41.31% reduction in iterations. Model behavior was further interpreted using feature importance analysis, permutation importance, and feature ablation studies to examine the contribution of individual features to iteration prediction. Statistical robustness is established using bootstrap confidence intervals, the Friedman test, and Holm-corrected Wilcoxon signed-rank tests. Full article

Small coastal crafts are often equipped with stock propellers selected from limited commercial options, without detailed matching among the hull, main engine, reduction gear, and propeller. This study presents a practical and integrated procedure for evaluating propulsion performance and designing a propeller tailored to a G/T 4.99 coastal craft under the fully loaded departure condition. The CFD resistance simulations were validated against model test measurements of resistance, sinkage, and trim and were applied over a wide range of vessel speeds. Self-propulsion simulations were subsequently conducted to evaluate the propulsive characteristics of the target vessel. Based on the evaluated self-propulsion characteristics, actual engine specifications and propeller installation data from Korean coastal vessels were used to select the main engine and reduction gear and to define practical ranges for the propeller diameter and pitch ratio. An initial propeller was determined using the $B_p-\delta$ method, after which the mean pitch ratio was iteratively corrected through full-scale performance prediction. The final propeller was selected by matching the predicted rotational speed under the trial condition to the target value determined by the engine and reduction gear while confirming the attainable vessel speed under the service condition with a 15% sea margin. The final propeller was evaluated through CFD self-propulsion simulations and compared with the stock propeller. At 16 and 18 knots, the designed propeller increased the overall propulsive efficiency by 8.4% and 10.1%, respectively, while reducing the required delivered power by 7.8% and 9.2%. These results demonstrate that the proposed procedure can improve the matching among the hull, machinery, and propeller and provide a practical basis for designing efficient propellers for small coastal crafts. Full article